Re: talk.origins faq Hitler claim part 3

nando_ronteltap@xxxxxxxxx wrote:
> Hershey wrote:
> >Syamsu wrote:
> >> So since you take account of beneficial being destroyed *within*
> >> natural selection, then I can say that natural selection results in
> >> individual beneficial variants being destroyed,
> >
> >No.
>
> Why don't you go read that paper, gee.

Which one?

The Two-Level Structure of Causation in Neo-Darwinian Theories of
Evolution (Mohan Matthen and Andre Ariew), in preparation

Mohan Matthen and Andre Ariew, "How to Understand Causal Relations in
Natural Selection: A Reply to Rosenberg and Bouchard" Biology and
Philosophy (forthcoming)

"Two Ways of Thinking About Fitness and Natural Selection" (co-author
Mohan Matthen), Journal of Philosophy, February 2002, Volume 49, No. 2,
pp. 53-83.

I wasn't able to access the last one and the others were not available
either.

> You are doing everything Ariew
> and Matthen say is a no-no in talking about probabilities, using many
> of the exactsame way of saying things which they point out as false.

Exactly what am I doing that is a no-no in talking about probabilities.
If you cannot explain that well enough for me to understand it, that
means that you don't understand it yourself. And you certainly have
given me no reason to believe that you know diddlysquat about
probability or statistics or anything related to natural selection.

> Aside from that, I prefer to make this issue simple. Since you say that
> beneficial destroyed is not part of natural selection at all, it
> therefore follows that chance of reproduction is no part of natural
> selection at all, for a chance of reproduction would have beneficial
> both reproducing and not reproducing, as the case may be.

Natural selection is about *differential* reproduction, or, more
accurately, the differential rates of failure to reproduce, since, as
Malthus points out, the capacity to reproduce always (eventually)
outweighs the carrying capacity of the environment.

Take my quite realistic (but simplified) example of insects, which has
an adult reproductive population of 1000 individual (500 mating pairs).
Each female produces 200 fertilized eggs each generation, meaning that
500 x 200 = 100,000 fertilized eggs are produced each generation.
Since the carrying capacity of the environment is such that only 1000
of the 100000 fertilized eggs (1%) will reach reproductive adulthood,
that means that 99% of the fertilized eggs will not contibute to the
next generation *no matter what* amount of differential reproductive
success goes on *within* this population.

Now, let's assume we are examining two variant phenotypes within this
population, and in generation 1 we have 75% showing dominant phenotype
A and 25% showing recessive phenotype a (this would be the case if 50%
of the alleles in the population are the dominant allele A and 50% are
recessive allele a for a single gene locus in H-W equilibrium).
According to simple Hardy-Weinberg expectations, if there were no
neutral drift and no natural selection, we would expect to see exactly
the same ratios in their progeny at both conception and in the 1% that
survive to reproduce each generation (I will assume that the carrying
capacity is constant, but it really doesn't matter).

However, because this is a real population of a real size rather than
an ideal population of infinite size, we are actually taking a random
sampling of alleles in producing the 100000 zygotes and taking a
sampling of the zygotes in reaching the 1000 lucky reproductive members
of Gen2 that will be the parents of Gen3. So I cannot expect, even in
the absence of selection, that there will be *exactly* 50% A and 50% a
alleles in the Gen2 reproductive adults. I will expect the numbers
(assuming no selective difference) to be *close* to 500 A and 500 a
(with 75% showing the dominant phenotype and 25% showing the recessive
phenotype). I can, of course, statistically determine whether the
results I get are reasonably *close* to 500 A and 500 a by simple
statistical tests. [Generally, if the result I get is one that is
within the range of results that I could expect by chance 95% of the
time from such a sampling, I would say that the result is
indistinguishable from 500::500.]

OTOH, if I observe about 500 reproducing individuals in G2 with
recessive phenotype a (instead of the H-W expected number of 250 (25%))
and see that, among the 500 survivors with phenotype A, that 2/3 are
heterozygotes (Aa) and 1/3 are homozygous A (AA), it would be very easy
to determine that this result was highly unlikely to be due to chance.
In fact, it would be quite clear that we are seeing many more than the
expected number of survivors who had the recessive phenotype surviving
to reproduction. That would be direct experimental evidence that there
was some *environmental* *causal* reason why I am seeing the
*selective* and *differential* excess survival of individuals with
phenotype a. [Note that this emphasizes that selection is for or
against phenotypes, not genotypes. There was no differential effect on
the survival of heterozygotes. The effect of selection on genotypes is
secondary to and a result of the effect on phenotypes. Also note that
I may not know what the environmental causal factor is that led to the
*differential reproductive success* of the recessive phenotype. I may
have suspects. And I certainly have reason to look for some
environmental cause. I would not have any reason to search for an
environmental variable that differentially affects individuals with
phenotype a or A if my statistical test said that the result is
indistinguishable from what could happen by chance alone.] This would
be a classic case of natural selection (the direct observation of
differential reproductive success). I have gone from the reproducing
population of Gen1 having 1000 A and 1000 a alleles to the same point
in Gen2 having 667 A and 1333 a alleles (more likely numbers
insignificantly different from these -- chance obviously does not cease
to exist because of selection).

But note that the vast majority of individuals conceived by the Gen1
parents that had the *selectively beneficial* recessive phenotype must
necessarily die because the carrying capacity only allows 1% of the
fertilized eggs to reach reproductive age. The difference is that only
98% of individuals with phenotype a died whereas 99.5% of those with
phenotype A died. That is significant *differential reproductive
success* in this context. In the absence of differential reproductive
success (aka natural selection) 99% of each phenotype would have died.

> Therefore statistics really have nothing whatsoever to do with your
> conception of natural selection, you just use statistics do calculate
> the environmental randomness, chance acting alone and whatnot, which
> you substract to arrive at pure natural selection which is a force that
> preserves beneficial and destroys injurous.

The above nonsense is exactly why you have given me no reason to
believe that you know diddlysquat about probability or statistics or
anything related to natural selection. Statistics is used just above
to *identify* whether or not an observation is likely to be, or
unlikely to be, due to natural selection -- differential reproductive
success due to some environmental cause or factor that differentially
discriminates between alternative phenotypes.

> So everytime the bee goes to the hallowed flower it is force, a cause
> acting, and everytime it goes to an injurous preserving it, it is
> chance acting not a cause.

I just gave an explicit example of a case where there is only 1%
survival to reproduction (because of environmental carrying capacity
constraints) in which one can easily and clearly either see no natural
selection (neutral drift) or can see natural selection (differential
reproductive success) occurring. The overall rate of survival has
nothing to do with whether natural selection occurs. If the overall
rate of survival were 10% rather than 1%, and other numbers adjusted
appropriately how would that affect or change the amount of or
direction (which phenotype is beneficial) of natural selection. If the
carrying capacity changed from generation to generation from 1% one
generation to 10% the next (say a dry year followed by a wet one), you
could easily still have a case where there is no differential
reproductive success between two variants or a case where there is
differential reproductive successs. It would still be possible to
identify the likelihood that natural selection occurred. You would use
the same methodology to determine whether there had or had not been
differential reproductive success. The statistical significance of the
*differential* rate of survival of different phenotypes is how you
determine the presence or absence of natural selection.

You are correct in one sense. There is no 'force' of natural
selection. Natural selection is all a question of the dumb, stupid,
unintelligent environment applying differential rates of death or
reduced fertility to populations. Natural selection has no mechanism
to *increase* the number of individuals with 'beneficial' phenotypes
beyond that which are initially conceived. Selection (by the
environment) only *differentially* affects the rate of losses from that
high point. It is losses and more losses all the way down till you
reach the reproductively successful. The environment only *favors*
(relatively) the reproduction of certain individuals, while continually
grinding away against all. It doesn't create or invent favored
individuals. It only differentially grinds away at different
categories of individuals. Of course, this is where the excess
Malthusian reproductive capacity of being able to produce more
individuals than could possibly survive or reproduce comes in handy.

> HAHAHAHA
>
> you understand nothing
>
> HAHAHAHAHA
>
> >Natural selection does
> >not and cannot prevent chance loss (meaning unrelated to the traits
> >being examined) from happening. It only means that one trait is
> >*favored* and has a greater *probability* of reproductive success than
> >the other.
>
> Sure your conception of natural selection is in terms of probability,
> it is two sides of two probabilities, the side of the beneficial
> variant being preserved, and the side of the injurous variant being
> eliminated. The other sides to these probabilities fall under the
> theory of reading dyslexia for basic statistics.

How so? The above nonsense is exactly why you have given me no reason
to believe that you know diddlysquat about probability or statistics or
anything related to natural selection. What the statistics I use
determine is whether or not the *difference* in relative reproductive
success (between the two possibilities) is significantly greater than
can be accounted for by chance. That is one probability measurement,
not two, much less four.
>
> regards,
> Mohammad Nor Syamsu

.

Relevant Pages

  • Re: How Our Brains Ignore Unpleasant Facts was: Re: The Reasonable
    ... success for an organism by acting on phenotype. ... Natural selection is the process by which favorable traits that are ... of reproducing organisms, and unfavorable traits that are heritable ... and reproduction between different genotypes, or even different genes, ...
    (talk.origins)
  • Re: How Our Brains Ignore Unpleasant Facts was: Re: The Reasonable
    ... success for an organism by acting on phenotype. ... Natural selection is the process by which favorable traits that are ... of reproducing organisms, and unfavorable traits that are heritable ... and reproduction between different genotypes, or even different genes, ...
    (talk.origins)
  • Re: How Our Brains Ignore Unpleasant Facts was: Re: The Reasonable
    ... success for an organism by acting on phenotype. ... Natural selection is the process by which favorable traits that are ... of reproducing organisms, and unfavorable traits that are heritable ... and reproduction between different genotypes, or even different genes, ...
    (talk.origins)
  • Re: How Our Brains Ignore Unpleasant Facts was: Re: The Reasonable
    ... > success for an organism by acting on phenotype. ... > Natural selection is the process by which favorable traits that are ... > of reproducing organisms, and unfavorable traits that are heritable ... > and reproduction between different genotypes, or even different genes, ...
    (talk.origins)
  • Re: How Our Brains Ignore Unpleasant Facts was: Re: The Reasonable
    ... success for an organism by acting on phenotype. ... Natural selection acts on the phenotype, ... and reproduction between different genotypes, or even different genes, ... For example, we can test "F = ma", and can falsify it. ...
    (talk.origins)