Complexity



friend <[email protected]> wrote:

>> JE:- Darwinian natural selection, exactly as Darwin
>> stated it but with his implicit assumptions made
>> explicit.

> BOH:- I think I should point out that John's definition of
> fitness excludes the possibility of drift (because he
> defines fitness in terms of the actual number of
> offspring, rather than the expected value).

F:- Is the expected value of any conceivable use?

JE:- Yes: when (which is most of the time) we do not have an
exact measure of total Darwinian fitness for every Darwinian
selectee within one population.

As long as the expected value:

1) Attempts to measure _total_ Darwinian fitness per
selectee and not just a sub total.

2) The expectation does not replace any actual, measured
Darwinian fitness total.

Then: the expected value is not misused. If the expected
value is just a simplified model of Darwinian total fitness
and attempts to compete and win against it, then such a
model of expected fitness is utterly misused because the
model seeks to replace the reality it is only attempting to
model! Unbelievably, this continues to be a general error
within gene centric Neo Darwinism.

> JE:- Dr O'Hara has misrepresented my position. Drift is
> _included_ as temporal variation (random variation over
> time) within Darwinian selective events. Mutation is
> included as random variation within genetic space.
> Darwinian evolution by natural selection has, from its
> inception, required variation that is assumed to be
> random. Only _non_ random forms of variation can validly
> compete against Darwinian selection for evolutionary
> causation. Sampling error is not one of them. Isn't it
> just obvious that an "expected value" is just an
> approximation of "the actual number of offspring" which I
> stress (yet again) is the TOTAL (requiring a time frame to
> complete) of only FERTILE forms reproduced into ONE
> population by EACH Darwinian selectee (one fertile form)?
> Please note that Dr O'Hara, refused to address the problem
> that I addressed in detail: setting up a _controlled_
> experiment to test if random patterns can _alone_, cause
> "evolution".

F:- I think only with genetic engineering is it possible to
even approach random patterns. And then it is very
imperfect practically. Very difficult.

JE:- Random patterns remain: random. Thus, they cannot be
predicted.

> AGAIN: Please state what would be EXCLUDED from such an
> _amazingly_ wide acceptance of what Dr Moran insists can
> _scientifically_ constitute "evolution", i.e. would Dr
> Moran please provide at least one example of a _non_
> evolutionary change within a biological system. PLEASE
> COMMENT ON THE BELOW:-
> 1) Drift defined as evolution is not testable because any
> random pattern can validly be assumed to be caused by
> either a random or non random process.
> 2) It is only possible to halt all selection within a
> natural population by forcing the _total_ number of
> _fertile_ forms reproduced by each Darwinian selectee
> in one population to remain equal. Does Dr Moran agree
> or disagree? Only by controlling selection can random
> genetic drift patterns even be observed acting _alone_.

G:- If then...

> Does Dr Moran dispute the fact that to just observe the
> effect of random patterns on their own, selection must be
> controlled? Does Dr Moran understand/teach the need for
> controlled experiments to his students? The predicted net
> result of drift without selection within a controlled
> experiment is the dissolution of every Darwinian selectee
> within that population. Does Dr Moran disagree with this
> prediction? If Dr Moran agrees with the prediction, how
> can he maintain that genetic dissolution that inevitably
> leads to a _lowering_ of Darwinian fitness for every
> selectee constitute "evolution"?

H:- I think it is much worse than this; my understanding may
be too narrow but too much has happened (eg. crossover)
before the individual is subject to very much selection to
regard it as 'just' random. So even with selection the
prognosis may not be so good.

JE:- My main point is: only NON random patterns can be
controlled,
H.a. eliminated. Random patterns cannot be eliminated, only
allowed for.

To test if ANY random pattern can cause evolution, selection
must be controlled and eliminated within an experimental
situation. Such a requirement is just _basic_ science. Dr
Moran et al wishes to classify just a random pattern as
"evolution" without having to supply any experimental proof
that ANY random pattern can produce evolution without
selection, i.e. he wishes to dictate what evolution is just
to suit his own purposes. What are his purposes? You will
have to ask Dr Moran that question...

> JE:- Genetic drift without selection cannot cause
> evolution even it can cause gene freq changes in a deme.
> Selection can cause evolution without genetic drift but
> genetic drift cannot cause evolution without selection.
> Random genetic drift is just temporal variation; nothing
> more and nothing less. Evolution at the gene centric level
> can only be validly defined as any non random gene freq.
> change in a deme.

I:- Is selection on the random change exactly *that* non
random change?

JE:- Yes, the selective event is NON random even if the
variation that is provided was only random. Ever since its
inception non random selection has required random
variation.

Regards,

John Edser Independent Researcher

PO Box 266 Church Pt NSW 2105 Australia

[email protected]
 
friend wrote:
> John Edser wrote:
>
>>
>>>JE:- Darwinian natural selection, exactly as Darwin
>>>stated it but with his implicit assumptions made
>>>explicit.
>>
>>BOH:- I think I should point out that John's definition of
>>fitness excludes the possibility of drift (because he
>>defines fitness in terms of the actual number of
>>offspring, rather than the expected value).
>>
>
>
> Is the expected value of any conceivable use?
>
Yes - it's predicted from models of selection, so you can
develop a determistic model for selection, and it will
usually give the same value as the expected value here.

Bob

--
Bob O'Hara

Dept. of Mathematics and Statistics
P.O. Box 4 (Yliopistonkatu 5) FIN-00014 University of
Helsinki Finland Telephone: +358-9-191 23743 Mobile:
+358 50 599 0540 Fax: +358-9-191 22 779 WWW:
http://www.RNI.Helsinki.FI/~boh/ Journal of Negative
Results - EEB: http://www.jnr-eeb.org
 
JE: Yes, the selective event is NON random even if the
variation that is provided was only random. Ever since its
inception non random selection has required random
variation.

MR:You're basic argument seems to be that randomness can't
occur without selection and vice versa. Its my
understanding most biologists accept random genetic drift
and natural selection not being necessarily mutually
exclusive. Your assertion would seem to be they are never
mutually exclusive. "F" remarks, " I think only with
genetic engineering is it possible to even approach
random patterns. And then it is very imperfect
practically. Very difficult. You reply, "Random patterns
remain: random.Thus, they cannot be predicted." Yet if
selection can't occur without randomness there must be a
way (perhaps still unknown) to predict how the randomness
of evolution translates into selection. If that is the
case then possibly one day it will be possible to predict
random patterns. I would think genetic engineering has
offered such opportunities.

"F" remarks "I think it is much worse than this; my
understanding may be too narrow but too much has happened
(eg. crossover) before the individual is subject to very
much selection to regard it as 'just' random. So even with
selection the prognosis may not be so good. I don't fully
understand meiosis but apparently the point here is crossing
over plays an important role in terms of "selection" for it
just to be 'random'.

You state, "My main point is: only NON random patterns can
be controlled, i.e. eliminated. Random patterns cannot be
eliminated, only allowed for." Certainly with genetic
engineering random patterns can be eliminated as well as
allowed for. You further state, "To test if ANY random
pattern can cause evolution, selection must be controlled
and eliminated within an experimental situation. Yet random
genetic drift can't be proved because it is impossible to
establish it has occurred in the absense of selection..

Michael Ragland
 
John Edser wrote:
>>>>JE:- Pardon me butting in but.. please state what would
>>>>be EXCLUDED from such an _amazingly_ wide acceptance of
>>>>what you insist can _scientifically_ constitute
>>>>"evolution", i.e. please provide at least one example of
>>>>a _non_ evolutionary change within a biological system.
>>>
>
>>BOH:- Any change that isn't heritable.
>
>
>>JE:- The above becomes a self fulfilling prophecy when
>>genetic epistasis is _defined_ as "inherited" but not
>>"heritable" and thus, "selectable".
>
>
> BOH:- Huh? From the OED definition of heritable: Naturally
> transmissible or transmitted from parent to offspring;
> hereditary. You're mixing up being heritable with the
> qunatitative genetic concept of "heritability" (=the
> proportion of variance in a trait due to additive genetic
> variation). They are not the same thing.
>
> JE:- "Heritability" and "heritable" mean the same thing: a
> selectable trait. You used the word "heritable" and not
> the word "heritability" as in: "Any change that isn't
> heritable".
>
In genetics and evolutionary biology, they don't mean the
same thing. Please try and educate yourself first before
making comments like this.

As for epistasis, if something isn't heritable, then it
can't be inherited. Genetic epistatic effects can be
inherited (because groupds of genes can be inherited
together), and hence does not come under what I was
suggesting.

<snip>

>>>>LM:- Did you have something else in mind that would
>>>> shift random genetic drift into second place?
>>>
>
>>>>JE:- Darwinian natural selection, exactly as Darwin
>>>>stated it but with his implicit assumptions made
>>>>explicit.
>>>
>
>>>BOH:- I think I should point out that John's definition
>>>of fitness excludes the possibility of drift (because he
>>>defines fitness in terms of the actual number of
>>>offspring, rather than the expected value).
>>
>
>>>JE:- Dr O'Hara has misrepresented my position. Drift is
>>>_included_ as temporal variation (random variation over
>>>time) within Darwinian selective events.
>>
>
>>BOH:- John, what is your definition of fitness? I was
>>specifically describing your definition of fitness, but in
>>your reply you didn't make any mention of it, so it's not
>>clear to me how I've misrepresented you.
>
>
>>JE:- It is not how I define it but how Darwin would have
>>defined it after his implicit assumptions were made
>>explicit. I have posted what Darwinian fitness is (and the
>>reverse engineering experiment needed to prove it)
>>countless times, including, within this thread.
>>_____________________________________________________
>>Darwinian fitness is the _total_ number of _fertile_ forms
>>reproduced by _one_ Darwinian selectee (one fertile form)
>>within _one_ population.
>>_____________________________________________________
>
>
> BOH:- OK, so my point is that drift is the difference
> between observed and [mathematical] expectation of the
> change in allele frequency (the expectation coming from a
> model of selection, using the conventional definition of
> fitness):
>
> JE:- The "the conventional definition of fitness" is only
> relative at just ONE single point in time but the
> Darwinian definition is an absolute total over a defined
> TIME FRAME:

The time frame is irrelevant to my argument, and discussing
it here will just make things even more confused.

<snip>
>
> BOH:- the total number of fertile offspring is the sum of
> the expected number, and the drift effect.
>
> JE:- Because any absolute measure of fitness has to, since
> it must also include all variation within one evolutionary
> change. Please note that a relative measure of fitness is
> not sufficient simply because you cannot differentiate
> between organism fitness altruism (OFA) and organism
> fitness mutualism (OFM) using just one relative
> comparative measure.
>
> While it is possible to separately measure the variation
> caused by random patterns from variation caused by non
> random patterns, it is not possible to say that a random
> variation pattern must have been caused by just, a random
> process. Drift theorists refuse to admit that this is the
> case and has always been the case throughout the history
> of science. This is why science throws out random patterns
> as inconclusive.
>
> BOH:- Your definition conflates the effects of selection
> and drift, by defining fitness in terms of the obseved
> number. Any random variation in the number of offspring
> that we would ascribe to drift you would include in the
> fitness measure, and hence would ascribe to selection.
>
> JE:- Yes,

OK, good. This is the point I was trying to make: you define
fitness in such a way as to include any drift effects into
it. So, for you drift does not exist, as it's in your
fitness measure.

Bob

--
Bob O'Hara

Dept. of Mathematics and Statistics
P.O. Box 4 (Yliopistonkatu 5) FIN-00014 University of
Helsinki Finland Telephone: +358-9-191 23743 Mobile:
+358 50 599 0540 Fax: +358-9-191 22 779 WWW:
http://www.RNI.Helsinki.FI/~boh/ Journal of Negative
Results - EEB: http://www.jnr-eeb.org
 
"Anon." <[email protected]> wrote in message
news:[email protected]...
> As for epistasis, if something isn't heritable, then it
> can't be inherited. Genetic epistatic effects can be
> inherited (because groups of genes can be inherited
> together), and hence does not come under what I was
> suggesting.

Pardon me for butting in, but some time ago I told John that
the reason that non-additive epistatic effects are usually
omitted from evolutionary models is that such effects are
not effectively heritable. Was this incorrect?

In saying this, I had in mind the situation in which the two
loci are on different chromosomes. I (and John, I think)
realize that there is some heritability when the loci are in
the same linkage group, but it seems to me that weak linkage
cannot be a factor in long term evolution. Strong linkage,
of course, might be a factor, but this doesn't argue against
the adequacy of the models because a strongly linked cluster
of loci can be treated as a single "super-gene" in the
additive models.
 
>>>JE:- Darwinian natural selection, exactly as Darwin
>>>stated it but with his implicit assumptions made
>>>explicit.

>>BOH:- I think I should point out that John's definition of
>>fitness excludes the possibility of drift (because he
>>defines fitness in terms of the actual number of
>>offspring, rather than the expected value).

> F:- Is the expected value of any conceivable use?

BOH:- Yes - it's predicted from models of selection, so you
can develop a determistic model for selection, and it will
usually give the same value as the expected value here.

JE:- Dr O'Hara has failed to say from which testable
theory/theories (views that can be verified or refuted)
these _non_ testable models are just simplifications/ over
simplifications, from. Unless Dr O'Hara specifically
nominates his missing theory/theories, he has misused
these models.

Best Wishes,

John Edser Independent Researcher

PO Box 266 Church Pt NSW 2105 Australia

[email protected]
 
JE: Yes, the selective event is NON random even if the
variation that is provided was only random. Ever since its
inception non random selection has required random
variation.

MR:You're basic argument seems to be that randomness can't
occur without selection and vice versa.

JE:- Some random variation is required if selection is to be
able operate. This does not need to be genetic drift because
this is not the only random pattern that can provide
variation. Also, it is impossible to stop all random
variation but it is possible to stop natural selection,
which provides a non random pattern.

MS:- Its my understanding most biologists accept random
genetic drift and natural selection not being necessarily
mutually exclusive.

JE:- All random processes are self exclsuive to all non
random processes because they are self contradictory.
However, just the observation of a random _pattern_ does not
exclude the possibility that a non random _process_ formed
it because they are not contradictory. A non random process,
can and does, form random patterns a lot/most of the time.

MT:- Your assertion would seem to be they are never mutually
exclusive.

JE:- My assertion was:- All defined non random processes are
mutually exclusive to all defined random processes.

MU:- "F" remarks, " I think only with genetic engineering is
it possible to even approach random patterns. And then it
is very imperfect practically. Very difficult. You reply,
"Random patterns remain: random.Thus, they cannot be
predicted." Yet if selection can't occur without
randomness there must be a way (perhaps still unknown) to
predict how the randomness of evolution translates into
selection.

JE:- The randomness of variation translates into evolution
if and only if, random variation is _selected_. This is the
only way known to include any defined random process within
a theory (can be varified or refuted) of evolution.

MV:- If that is the case then possibly one day it will be
possible to predict random patterns. I would think
genetic engineering has offered such opportunities.

JE:- Ther only possible way to predict random patterns is
show they were not random patterns in the first place.

MW:- "F" remarks "I think it is much worse than this; my
understanding may be too narrow but too much has happened
(eg. crossover) before the individual is subject to very
much selection to regard it as 'just' random. So even
with selection the prognosis may not be so good. I don't
fully understand meiosis but apparently the point here is
crossing over plays an important role in terms of
"selection" for it just to be 'random'.

You state, "My main point is: only NON random patterns can
be controlled, i.e. eliminated. Random patterns cannot be
eliminated, only allowed for." Certainly with genetic
engineering random patterns can be eliminated as well as
allowed for.

JE:- Random patterns cannot be entirely eliminated because
BOTH random and non random processes commonly cause them.
All you can do is allow for a random componant. This is most
commonly allowed for as a defined zero observation, e.g. a
defined random distribution such as the HW binomal
distribution model of allels in a heuristic infinite
population.

MX:- You further state, "To test if ANY random pattern can
cause evolution, selection must be controlled and
eliminated within an experimental situation. Yet random
genetic drift can't be proved because it is impossible to
establish it has occurred in the absense of selection..

JE:- You appear to have confused a random _pattern_ with a
random _process_. Don't worry, Neo Darwinists do so almost
all of the time. Note that this does not make such an
error, correct! Yes, "genetic drift can't be proved because
it is impossible to establish it has occurred in the
absence of selection" because genetic drift is a supposed
random PROCESS and not just a supposed random PATTERN.
However, it is possible to allow for only a random pattern
to cause all gene freq. changes within one natural
population but even here, it is not possible to suggest
that only a random process (genetic drift) caused this
controlled, random pattern.

It is possible to control for selection (eliminate the non
random pattern caused by Darwinian selection theory within a
real experimental situation) but it is not possible to
control for the process of random genetic sampling error
(genetic drift), only allow for it, because genetic drift
cannot be eliminated from any natural population.

Regards

John Edser Independent Researcher

PO Box 266 Church Pt NSW 2105

Australia

[email protected]
 
>>>>JE:- Pardon me butting in but.. please state what would
>>>>be EXCLUDED from such an _amazingly_ wide acceptance of
>>>>what you insist can _scientifically_ constitute
>>>>"evolution", i.e. please provide at least one example of
>>>>a _non_ evolutionary change within a biological system.

>>BOH:- Any change that isn't heritable.

>>JE:- The above becomes a self fulfilling prophecy when
>>genetic epistasis is _defined_ as "inherited" but not
>>"heritable" and thus, "selectable".

> BOH:- Huh? From the OED definition of heritable: Naturally
> transmissible or transmitted from parent to offspring;
> hereditary. You're mixing up being heritable with the
> qunatitative genetic concept of "heritability" (=the
> proportion of variance in a trait due to additive genetic
> variation). They are not the same thing.

> JE:- "Heritability" and "heritable" mean the same thing: a
> selectable trait. You used the word "heritable" and not
> the word "heritability" as in: "Any change that isn't
> heritable".

BOH:- In genetics and evolutionary biology, they don't mean
the same thing. Please try and educate yourself first before
making comments like this.

JE:- EITHER, something is defined as selectable or it is not
so defined. If, according to you, "heritable" and
"heritablility" don't mean the same thing, please explain
the difference.

I was attempting to "try and educate" myself re: the common
use of these concepts within Neo Darwinism. However, what I
am being taught makes no rational sense. All I can make out
is that quantitative genetics _models_ of “heritability”
delete non additive variation as "heritable" and thus
"selectable" information even though it is actually
"inherited". These models alone, dictate that gene fitness
epistasis, which ties all gene fitnesses together at just
_one_ level of selection, can be validly deleted via the
simplification. This simplification and nothing else,
allowed Hamilton's heuristic model of independent gene
fitnesses (only allow additive variation as heritable
information) to be able to compete and win against Darwinian
theory forcing organism fitness altruism (OFA) which is
totally excluded by Darwinism.

AGAIN: Please supply a real biological example (not just a
generalisation) of:

1) Something “inherited” in nature.
2) Something “heritable” in nature.
3) The “heritability” of something in nature.

BOH:- As for epistasis, if something isn't heritable, then
it can't be inherited.

JE:- You were suggesting that additive variation was
"heritable" and thus "selectable" information but the non
additive variation was "non heritable" and thus "non
selectable".

So, genetic epistasis is actually "inherited" but it is not
"heritable" so it is not "selectable" when it is composed
of non additive variation within quantitative models? This
means to me that you are suggesting that the critical
relationships between genomic genes that are non additive
(constitute non additive variation), which however
constitutes the bulk or "inherited" information that forces
all genomic genes to be selected at exactly the same,
single level of selection: the Darwinian fertile organism
level, and therefore constitute the major componant of
Darwinian fitness, do not constitute a heritable and thus
selectable fitness?

BOH:- Genetic epistatic effects can be inherited (because
groupds of genes can be inherited together), and hence does
not come under what I was suggesting.

JE:- We know and agree that "epistatic effects can be
inherited" as a testable theory of nature. What is under
dispute is the use or misuse of quantitative genetics models
of this theory which delete gene fitness epistasis as
"heritable" and thus "selectable" information as just a
simplification and then allow such simplified models to
compete and win against the theory they were simplified
from, e.g. Hamilton’s model.

What is they theory, what is the model, and which has been
derived from what by what means?

>>BOH:- John, what is your definition of fitness? I was
>>specifically describing your definition of fitness, but in
>>your reply you didn't make any mention of it, so it's not
>>clear to me how I've misrepresented you.

>>JE:- It is not how I define it but how Darwin would have
>>defined it after his implicit assumptions were made
>>explicit. I have posted what Darwinian fitness is (and the
>>reverse engineering experiment needed to prove it)
>>countless times, including, within this thread.
>>_____________________________________________________
>>Darwinian fitness is the _total_ number of _fertile_ forms
>>reproduced by _one_ Darwinian selectee (one fertile form)
>>within _one_ population.
>>_____________________________________________________

> BOH:- OK, so my point is that drift is the difference
> between observed and [mathematical] expectation of the
> change in allele frequency (the expectation coming from a
> model of selection, using the conventional definition of
> fitness):

> JE:- The "the conventional definition of fitness" is only
> relative at just ONE single point in time but the
> Darwinian definition is an absolute total over a defined
> TIME FRAME:

BOH:- The time frame is irrelevant to my argument, and
discussing it here will just make things even more confused.

JE:- “The time frame is irrelevant” to your argument?
Including it will make things MORE confused? Any argument
about fitness that just deletes a Darwinian fitness TOTAL
such as Hamiltons is just, hopelessly confused.

Why do you steadfastly refuse to comment on the experimental
proof I supplied that that Darwinian fitness is

4) exactly as I defined it.
5) the only process that can can halt all selection when
reversed engineered within a natural population and not
just a simulated model.
6) Makes redundant all multi levels of selection, including
Hamiltons.

<snip>
>
> BOH:- the total number of fertile offspring is the sum of
> the expected number, and the drift effect.

> JE:- Because any absolute measure of fitness has to, since
> it must also include all variation within one evolutionary
> change. Please note that a relative measure of fitness is
> not sufficient simply because you cannot differentiate
> between organism fitness altruism (OFA) and organism
> fitness mutualism (OFM) using just one relative
> comparative measure. While it is possible to separately
> measure the variation caused by random patterns from
> variation caused by non random patterns, it is not
> possible to say that a random variation pattern must have
> been caused by just, a random process. Drift theorists
> refuse to admit that this is the case and has always been
> the case throughout the history of science. This is why
> science throws out random patterns as inconclusive.
>
> BOH:- Your definition conflates the effects of selection
> and drift, by defining fitness in terms of the obseved
> number. Any random variation in the number of offspring
> that we would ascribe to drift you would include in the
> fitness measure, and hence would ascribe to selection.

> JE:- Yes,

BOH:- OK, good. This is the point I was trying to make: you
define fitness in such a way as to include any drift effects
into it. So, for you drift does not exist, as it's in your
fitness measure.

JE:- Your conclusion: “for you drift does not exist, as it's
in your fitness measure” must separate drift as just a
_pattern_ from drift as a _process_. Drift as a real random
process, is never proven but drift as just a random pattern
has to be assumed. This is because random and non random
patterns must be able to be separated using statistical
tests otherwise the unique non ranndom pattern that
selection causes cannot be varified/refuted in nature.

Regards,

John Edser Independent Researcher

PO Box 266 Church Pt NSW 2105 Australia

[email protected]
 
Perplexed in Peoria wrote:
> "Anon." <[email protected]> wrote in
> message news:[email protected]...
>
>>As for epistasis, if something isn't heritable, then it
>>can't be inherited. Genetic epistatic effects can be
>>inherited (because groups of genes can be inherited
>>together), and hence does not come under what I was
>>suggesting.
>
>
> Pardon me for butting in, but some time ago I told John
> that the reason that non-additive epistatic effects are
> usually omitted from evolutionary models is that such
> effects are not effectively heritable. Was this incorrect?
>
Yes this was - the effects can be inherited, but the
strength of directional selection is not affected by the non-
additive epistatic effects.

This will be explained in Vol. 2 of Lynch & Walsh, if it is
ever published. But some of the material is here:
<http://nitro.biosci.arizona.edu/zbook/volume_2/vol2.html>

> In saying this, I had in mind the situation in which the
> two loci are on different chromosomes. I (and John, I
> think) realize that there is some heritability when the
> loci are in the same linkage group, but it seems to me
> that weak linkage cannot be a factor in long term
> evolution. Strong linkage, of course, might be a factor,
> but this doesn't argue against the adequacy of the models
> because a strongly linked cluster of loci can be treated
> as a single "super-gene" in the additive models.
>
The additive by additive component of variation does have an
effect on selection, but it is transient. See Chapter 4 of
Lynch & Walsh on the web page above.

Bob

--
Bob O'Hara

Dept. of Mathematics and Statistics
P.O. Box 4 (Yliopistonkatu 5) FIN-00014 University of
Helsinki Finland Telephone: +358-9-191 23743 Mobile:
+358 50 599 0540 Fax: +358-9-191 22 779 WWW:
http://www.RNI.Helsinki.FI/~boh/ Journal of Negative
Results - EEB: http://www.jnr-eeb.org
 
in article [email protected], Anon. at
[email protected] wrote on 6/19/04 3:40 PM:

> Perplexed in Peoria wrote:
>> "Anon." <[email protected]> wrote in
>> message news:[email protected]...
>>
>>> As for epistasis, if something isn't heritable, then it
>>> can't be inherited. Genetic epistatic effects can be
>>> inherited (because groups of genes can be inherited
>>> together), and hence does not come under what I was
>>> suggesting.
>>
>>
>> Pardon me for butting in, but some time ago I told John
>> that the reason that non-additive epistatic effects are
>> usually omitted from evolutionary models is that such
>> effects are not effectively heritable. Was this
>> incorrect?
>>
> Yes this was - the effects can be inherited, but the
> strength of directional selection is not affected by the
> non-additive epistatic effects.
>
> This will be explained in Vol. 2 of Lynch & Walsh, if it
> is ever published. But some of the material is here: <htt-
> p://nitro.biosci.arizona.edu/zbook/volume_2/vol2.html>
>
>> In saying this, I had in mind the situation in which the
>> two loci are on different chromosomes. I (and John, I
>> think) realize that there is some heritability when the
>> loci are in the same linkage group, but it seems to me
>> that weak linkage cannot be a factor in long term
>> evolution. Strong linkage, of course, might be a factor,
>> but this doesn't argue against the adequacy of the models
>> because a strongly linked cluster of loci can be treated
>> as a single "super-gene" in the additive models.
>>
> The additive by additive component of variation does have
> an effect on selection, but it is transient. See Chapter 4
> of Lynch & Walsh on the web page above.

An important limitation to the generality of this body of
theory, as explained by Dr. O'Hara, is that it usually
assumes no linkage among genes. We know that this assumption
is generally false, and I think we have yet to come to grips
with the implications of linkage, and the potential for
inheritance of coadapted gene complexes (functionally
coherent linkage groups), as it relates to patterns of
phenotypic development and natural selection.

Cheers,

Guy
 
John Edser wrote:
>>>>>JE:- Pardon me butting in but.. please state what would
>>>>>be EXCLUDED from such an _amazingly_ wide acceptance of
>>>>>what you insist can _scientifically_ constitute
>>>>>"evolution", i.e. please provide at least one example
>>>>>of a _non_ evolutionary change within a biological
>>>>>system.
>>>>
>
>>>BOH:- Any change that isn't heritable.
>>
>
>>>JE:- The above becomes a self fulfilling prophecy when
>>>genetic epistasis is _defined_ as "inherited" but not
>>>"heritable" and thus, "selectable".
>>
>
>>BOH:- Huh? From the OED definition of heritable: Naturally
>>transmissible or transmitted from parent to offspring;
>>hereditary. You're mixing up being heritable with the
>>qunatitative genetic concept of "heritability" (=the
>>proportion of variance in a trait due to additive genetic
>>variation). They are not the same thing.
>
>
>>JE:- "Heritability" and "heritable" mean the same thing: a
>>selectable trait. You used the word "heritable" and not
>>the word "heritability" as in: "Any change that isn't
>>heritable".
>
>
> BOH:- In genetics and evolutionary biology, they don't
> mean the same thing. Please try and educate yourself first
> before making comments like this.
>
> JE:- EITHER, something is defined as selectable or it is
> not so defined. If, according to you, "heritable" and
> "heritablility" don't mean the same thing, please explain
> the difference.
>
> I was attempting to "try and educate" myself re: the
> common use of these concepts within Neo Darwinism.

If you're trying to educate yourself, then please do so by
asking about things were you are ignorant, rather than
making statement based on your ignorance.

However,
> what I am being taught makes no rational sense. All I can
> make out is that quantitative genetics _models_ of
> “heritability” delete non additive variation as
> "heritable" and thus "selectable" information even though
> it is actually "inherited".

I would suggest you read a text-book on the subject - to be
honest, I see little point in spending my time trying to
explain this to you, as past (and present) experience shows
me that you aren't able to follow an argument without
throwing in your own incorrect interpretations. To whit:

> JE:- You were suggesting that additive variation was
> "heritable" and thus "selectable" information but the non
> additive variation was "non heritable" and thus "non
> selectable".
>
NO! NO! NO! I have never suggested that non-additive
variation is non-heritable.

> So, genetic epistasis is actually "inherited" but it is
> not "heritable"

For F***s sake! Read what I write - it is heritable and I've
never claimed anything otherwise.

Please go away, and read the literature - try Falconer and
MacKay, for example, or Lynch and Walsh (search for
"quantiative genetics" on your favourite web-based books
store!). Once you have educated yourself about what
quantitative genetics does and does not say (and how it
defines specific terms, e.g. heritability), then you will be
in a position to discuss.

<snip>

Bob

--
Bob O'Hara

Dept. of Mathematics and Statistics
P.O. Box 4 (Yliopistonkatu 5) FIN-00014 University of
Helsinki Finland Telephone: +358-9-191 23743 Mobile:
+358 50 599 0540 Fax: +358-9-191 22 779 WWW:
http://www.RNI.Helsinki.FI/~boh/ Journal of Negative
Results - EEB: http://www.jnr-eeb.org
 
>>BOH:- Huh? From the OED definition of heritable: Naturally
>>transmissible or transmitted from parent to offspring;
>>hereditary. You're mixing up being heritable with the
>>qunatitative genetic concept of "heritability" (=the
>>proportion of variance in a trait due to additive genetic
>>variation). They are not the same thing.

>>JE:- "Heritability" and "heritable" mean the same thing: a
>>selectable trait. You used the word "heritable" and not
>>the word "heritability" as in: "Any change that isn't
>>heritable".

> BOH:- In genetics and evolutionary biology, they don't
> mean the same thing. Please try and educate yourself first
> before making comments like this.

> JE:- EITHER, something is defined as selectable or it is
> not so defined. If, according to you, "heritable" and
> "heritablility" don't mean the same thing, please explain
> the difference.

JE:- AGAIN: EITHER explicitly state they mean the same thing
OR explain the difference!

>snip<

> JE:- what I am being taught makes no rational sense. All I
> can make out is that quantitative genetics _models_ of
> “heritability” delete non additive variation as
> "heritable" and thus "selectable" information even though
> it is actually "inherited".

BOH:- I would suggest you read a text-book on the subject -
to be honest, I see little point in spending my time trying
to explain this to you, as past (and present) experience
shows me that you aren't able to follow an argument without
throwing in your own incorrect interpretations.

JE:- Either additive and non additive gene fitness
epistasis were both regarded as constituting heritable and
thus selectable information within population genetics
over simplified models, or they are not. Very clearly,
form the very inception of population genetics
assumptions, they were NOT.

>From their very inception, over simplified population
genetics models of biological theory separated additive
variation from non additive variation because these
mathematical models only regarded additive variation as
constituting a "heritable" and thus "selectable" information
event even though non additive variation was defined
_theoretically_ as, "inherited".

http://instruct.uwo.ca/zoology/441a/hist3.html

Quote form the above website: "..he [Fisher] laid much of
the foundation of modern biological statistics (e.g. he
recommended variance - s2 - squared deviations from the mean
- as a measure of variation rather than the standard
deviation, s, since variances are additive when their
sources are independent) Of direct relevance here, he
recognised not only that all variation can be partitioned
into genetic and non-genetic causes, but that, while the
correlation among siblings is an expression of additive
(selectable) genetic variance, there is also a genetic
residue, including the effects of linkage, dominance and
epistasis."

The vast bulk of inherited information: non additive
variation was relegated by Fisher to just "a genetic
residue" and there it has stayed. Clearly, non testable
population genetics models are just a vast over
simplification of testable biological theory,
i.e. the theory that anything inherited must be heritable
was over simplified within population genetics models
to mean anything inherited as just additive variation
is defined within these models, as heritable. Of course
these models have real worth. However this worth has
been terribly diminished via their consistent misuse.

>snip embarrassing BOH invective<

Yours,

John Edser Independent Researcher

PO Box 266 Church Pt NSW 2105 Australia

[email protected]
 
You must log in or register to reply here.
Top Bottom Back