Dawkins on Kimura



John Edser wrote:
> John Edser wrote:
>
>>>TH:- But can we go further and say 'what isn't genetic
>>>drift (in the most broad sense of the word)
>>
>>
>>BOH:- Well, we could use a diffusion model, in which drift
>>wouldn't be drift. Sorry, I couldn't resist. To translate,
>>selection would be the difference between the expected
>>allele frequencies at two times points.
>>
>>JE:- Typically BOH replaces testable reality with just his
>>"expectations" of it. In simple terms such an absurd event
>>is termed: propaganda. The master was Stalin. He turned
>>reality into his "expectations" and millions perished.
>
>
> BOH:- John, please read up on simple probability theory
> before you write another post ridiculing expectations.
> They have a precise mathematical definition, which you
> seem to be totally unaware of.
>
> As for testability, you might like to read up about a
> wonderful subject called "statistics". It's used to fit
> models to data, which is an esential component of testing
> hypotheses. One of the things you can do is to estimate
> expected values (e.g. of allele frequency changes).
>
> JE:- Once again BOH asks us to replace a testable reality
> with just his expectations of it.

<snip>

This is just a sematic game. The expectation is of a random
variable, not of reality. There's no point in continuing
this if you don't even understand the basics.

Go away, read and learn about simple probability theory, and
about statistical method, and then return to this.

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
 
Guy Hoelzer <[email protected]> wrote in
news:[email protected]:

> in article [email protected], William
> Morse at [email protected] wrote on 3/15/04 8:38 AM:

>> Given my previous attempt on this thread to differentiate
>> phenomena primarily attributable to selection from
>> phenomena primarily attributable to drift (all are in
>> fact products of both), I would request that you clarify
>> what you mean by "all of these phenomena".

> In the context of the discussion, I meant both minor
> distinctions between sibling species and "more complicated
> examples of evolution."

>> The English and Irish finches may meet the requirements
>> of a default assumption for drift, but they may not, and
>> drift only qualitifies as a default assumption for a
>> limited set of observed phenotypic differences.

> Hmmm. Can you name one example of an "observed phenotypic
> difference" for which drift would not qualify as a
> "default" explanation?

Let's see: the big ears of elephants, (which I have
previously mentioned);the long noses, sticky tongues, and
lack of teeth in the anteaters, pangolins, and spiny
anteaters;the fact that both cetaceans and birds have good
vocal abilities;the recently acquired resistance of
Staphylococcus aureus to antibiotics; sickle cell anemia in
humans; eusociality in ants, termites, and naked mole rats;
development of hooves in perissodactyla and artiodactyla;
homeothermy in birds, mammals, and beehives; large brain
size in humans; morphology of the head in deer and kangaroos
vs. leopards and hyenas;conspicuous colors in monarch
butterflies, poison dart frogs, and coral snakes; the
relative sizes of testicles and penises in humans,
chimpanzees, and gorillas; the eye of vertebrates and
cephalopods, transparency in tunicates and comb
jellies;morphology of caecilians and earthworms; the
relationship of skeletal mass vs. body mass in mammals from
mice to elephants:aggregation in dictyostelium, myxococcus
fulvus, and physarum viride - oh I forgot you only asked for
one example :)

Now you may object that most of the above are similarities,
not differences and I stand guilty as charged. But the
similarities are all between animals that are to greater or
lesser degree unrelated. The flip side of this is that they
are more different from more closely related species. And
"drift" doesn't explain this, but "selection" does (see my
earlier follow for why the terms are in quotes).

You can probably give an equally long litany of examples of
"drift". This is exactly my point. You argued in another
follow on this thread that neutral drift should always be
the null hypothesis. I disagree strongly with that
statement. Both drift and selection always occur, but their
influence differs in different circumstances. I think we
know enough about evolution to recognize when one or the
other is _likely_ (note emphasis) to be dominant. So we do
not need to posit one or the other as a universal default.

Yours,

Bill Morse
 
>>>TH:- But can we go further and say 'what isn't genetic
>>>drift (in the most broad sense of the word)

>>BOH:- Well, we could use a diffusion model, in which drift
>>wouldn't be drift. Sorry, I couldn't resist. To translate,
>>selection would be the difference between the expected
>>allele frequencies at two times points.

>>JE:- Typically BOH replaces testable reality with just his
>>"expectations" of it. In simple terms such an absurd event
>>is termed: propaganda. The master was Stalin. He turned
>>reality into his "expectations" and millions perished.

> BOH:- John, please read up on simple probability theory
> before you write another post ridiculing expectations.
> They have a precise mathematical definition, which you
> seem to be totally unaware of. As for testability, you
> might like to read up about a wonderful subject called
> "statistics". It's used to fit models to data, which is an
> esential component of testing hypotheses. One of the
> things you can do is to estimate expected values (e.g. of
> allele frequency changes).

> JE:- Once again BOH asks us to replace a testable reality
> with just his expectations of it.

<snip>

BOH:- This is just a sematic game.

JE:- No, it is your "game" of misrepresentation.
Typically, you just snip most of what I write, fail to
answer the propositions I discuss and then misrepresent
what I did write.

1) Estimates are only valid if they are of some _testable_
reality, reality is NOT just a proposed estimate of
something that you could never measure.

2) You cannot replace a testable theory of nature such as
Darwin's which defines an exact fitness per defined
selectee with just non testable Neo Darwinian "estimates"
of things which you cannot even define. Over simplified
mathematical models cannot replace testable theories of
biology, within any valid science of biology.

BOH:- The expectation is of a random variable, not of
reality.

JE:- Which makes your attempt to replace a testable
Darwinian reality with such an estimate, even more absurd.

Best Wishes,

John Edser Independent Researcher

PO Box 266 Church Pt NSW 2105 Australia

[email protected]
 
Jeffrey Turner <[email protected]> wrote

> > The minimal definition of evolution is "a change in the
> > heritable characteristices of a population over time."
> > This does not require speciation. The change can be due
> > to random genetic drift or natural selection or possibly
> > other mechanisms. All scientists agree that it's
> > possible to have evolution without speciation and
> > without adaptation.
>
> Yeah, well. Given that loose a definition, sure it is.
> Drift is ubiquitous. It just isn't very important in the
> real world. Be glad you've got college boards of trustees
> to pay your salaries.

<snip>

> > We shouldn't have to be defending random genetic drift
> > against those who would dismiss it as a mechanism of
> > evolution.
>
> Electrical circuits all have noise, too. In understanding
> the mechanisms we generally time average measurements so
> that noise doesn't get in the way of our understanding of
> fundamental principles. Go study your genetic noise, I
> won't interrupt your navel gazing again.
>
> > Those sorts of things should be moved to talk.origins or
> > some other newsgroup that's intended for people who
> > don't know much about evolution.
>
> Zounds, thou hast cut me to the quick. I know enough about
> evolution to realize that genetic drift doesn't explain
> anything about it. Drift happens, yawn.

Jeffrey, you're a breath of fresh air here in SBE.

Here's another approach you might take in opposition to
this nonsensical supposition that drift represents a form
of evolution distinct from natural selection. As you know
they claim that genetic drift is the result of random
causation. And if we follow the logic of this argument that
would mean that natural selection must be the result of
nonrandom causation. Ask them to substantiate this. You
won't get a response.

Jim
 
John Edser wrote:

>>>>TH:- But can we go further and say 'what isn't genetic
>>>>drift (in the most broad sense of the word)
>
>
>>>BOH:- Well, we could use a diffusion model, in which
>>>drift wouldn't be drift. Sorry, I couldn't resist. To
>>>translate, selection would be the difference between the
>>>expected allele frequencies at two times points.
>
>
>>>JE:- Typically BOH replaces testable reality with just
>>>his "expectations" of it. In simple terms such an absurd
>>>event is termed: propaganda. The master was Stalin. He
>>>turned reality into his "expectations" and millions
>>>perished.
>
>
>>BOH:- John, please read up on simple probability theory
>>before you write another post ridiculing expectations.
>>They have a precise mathematical definition, which you
>>seem to be totally unaware of. As for testability, you
>>might like to read up about a wonderful subject called
>>"statistics". It's used to fit models to data, which is an
>>esential component of testing hypotheses. One of the
>>things you can do is to estimate expected values (e.g. of
>>allele frequency changes).
>
>
>>JE:- Once again BOH asks us to replace a testable reality
>>with just his expectations of it.
>
>
> <snip>
>
> BOH:- This is just a sematic game.
>
> JE:- No, it is your "game" of misrepresentation.

No, your statement confuses two uses of the word
"expectation". I was using the word in a technical sense,
which I then tried to explain. Your senetence seemed to use
the non-technical sense of the word, which is a different
matter. Your sentence implies that in replacing the real
world with something else, I am not making any allowance for
the fact that the real world may be different. This is not
the case, and the implication in your sentence comes purely
from the different uses of a word. In that sense it is
purely a matter of playing with semantics.

<snip>
>
> BOH:- The expectation is of a random variable, not of
> reality.
>
> JE:- Which makes your attempt to replace a testable
> Darwinian reality with such an estimate, even more absurd.
>
The only "Darwinian reality" is the real world. In our
theories we have to replace the real world with something
simpler, otherwise we cannot understand it.

Bob

--
Bob O'Hara Department 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: www.jnr-eeb.org
 
On Mon, 22 Mar 2004 17:30:23 +0000 (UTC), "Anon."
<[email protected]> wrote:

>John Edser wrote:
>
>>>>>TH:- But can we go further and say 'what isn't genetic
>>>>>drift (in the most broad sense of the word)
>>
>>
>>>>BOH:- Well, we could use a diffusion model, in which
>>>>drift wouldn't be drift. Sorry, I couldn't resist. To
>>>>translate, selection would be the difference between the
>>>>expected allele frequencies at two times points.
>>
>>
>>>>JE:- Typically BOH replaces testable reality with just
>>>>his "expectations" of it. In simple terms such an absurd
>>>>event is termed: propaganda. The master was Stalin. He
>>>>turned reality into his "expectations" and millions
>>>>perished.
>>
>>
>>>BOH:- John, please read up on simple probability theory
>>>before you write another post ridiculing expectations.
>>>They have a precise mathematical definition, which you
>>>seem to be totally unaware of. As for testability, you
>>>might like to read up about a wonderful subject called
>>>"statistics". It's used to fit models to data, which is
>>>an esential component of testing hypotheses. One of the
>>>things you can do is to estimate expected values (e.g. of
>>>allele frequency changes).
>>
>>
>>>JE:- Once again BOH asks us to replace a testable reality
>>>with just his expectations of it.
>>
>>
>> <snip>
>>
>> BOH:- This is just a sematic game.
>>
>> JE:- No, it is your "game" of misrepresentation.
>
>No, your statement confuses two uses of the word
>"expectation". I was using the word in a technical sense,
>which I then tried to explain. Your senetence seemed to use
>the non-technical sense of the word, which is a different
>matter. Your sentence implies that in replacing the real
>world with something else, I am not making any allowance
>for the fact that the real world may be different. This is
>not the case, and the implication in your sentence comes
>purely from the different uses of a word. In that sense it
>is purely a matter of playing with semantics.
>
><snip>
>>
>> BOH:- The expectation is of a random variable, not of
>> reality.
>>
>> JE:- Which makes your attempt to replace a testable
>> Darwinian reality with such an estimate, even more
>> absurd.
>>
>The only "Darwinian reality" is the real world. In our
>theories we have to replace the real world with something
>simpler, otherwise we cannot understand it.
>
Bob, John Edser simply refuses to accept as scientific
anything that involves probability or probabilistic
phenomena or processes. I don't know what he does with
quantum theory or analysis of membrane noise or diffusion or
anything related. It doesn't do any good arguing.

If you think this is strange, you should read his stuff on
set theory!
 
in article [email protected], William Morse at
[email protected] wrote on 3/19/04 7:51 AM:

> Guy Hoelzer <[email protected]> wrote in
> news:[email protected]:
>
>> in article [email protected], William
>> Morse at [email protected] wrote on 3/15/04 8:38 AM:
>
>
>>> Given my previous attempt on this thread to
>>> differentiate phenomena primarily attributable to
>>> selection from phenomena primarily attributable to drift
>>> (all are in fact products of both), I would request that
>>> you clarify what you mean by "all of these phenomena".
>
>> In the context of the discussion, I meant both minor
>> distinctions between sibling species and "more
>> complicated examples of evolution."
>
>>> The English and Irish finches may meet the requirements
>>> of a default assumption for drift, but they may not, and
>>> drift only qualitifies as a default assumption for a
>>> limited set of observed phenotypic differences.
>
>> Hmmm. Can you name one example of an "observed phenotypic
>> difference" for which drift would not qualify as a
>> "default" explanation?
>
> Let's see: the big ears of elephants, (which I have
> previously mentioned);the long noses, sticky tongues, and
> lack of teeth in the anteaters, pangolins, and spiny
> anteaters;the fact that both cetaceans and birds have good
> vocal abilities;the recently acquired resistance of
> Staphylococcus aureus to antibiotics; sickle cell anemia
> in humans; eusociality in ants, termites, and naked mole
> rats; development of hooves in perissodactyla and
> artiodactyla; homeothermy in birds, mammals, and beehives;
> large brain size in humans; morphology of the head in deer
> and kangaroos vs. leopards and hyenas;conspicuous colors
> in monarch butterflies, poison dart frogs, and coral
> snakes; the relative sizes of testicles and penises in
> humans, chimpanzees, and gorillas; the eye of vertebrates
> and cephalopods, transparency in tunicates and comb
> jellies;morphology of caecilians and earthworms; the
> relationship of skeletal mass vs. body mass in mammals
> from mice to elephants:aggregation in dictyostelium,
> myxococcus fulvus, and physarum viride - oh I forgot you
> only asked for one example :)
>
> Now you may object that most of the above are
> similarities, not differences and I stand guilty as
> charged.

I'll let it slide this time.

> But the similarities are all between animals that are to
> greater or lesser degree unrelated. The flip side of this
> is that they are more different from more closely related
> species. And "drift" doesn't explain this, but "selection"
> does (see my earlier follow for why the terms are in
> quotes).

I agree with you that convergent evolution smacks of
selection and adaptation, but such examples do not address
my challenge, which was about the default explanation for
particular traits in particular taxa. Why, for example,
would you think that elephant ears alone were more likely to
be the result of selection than some other trait?

I will put off for now a discussion of non-adaptive and non-selection-
based mechanisms of convergence, which I think should not be
ignored as much as they have been.

> You can probably give an equally long litany of examples
> of "drift". This is exactly my point. You argued in
> another follow on this thread that neutral drift should
> always be the null hypothesis. I disagree strongly with
> that statement. Both drift and selection always occur, but
> their influence differs in different circumstances. I
> think we know enough about evolution to recognize when one
> or the other is _likely_ (note emphasis) to be dominant.
> So we do not need to posit one or the other as a universal
> default.

I think we agree more than disagree here. My comment about
the default or null explanation applies only to the
context in which drift and selection are assumed to be
independent "forces."

Cheers,

Guy
 
Jim McGinn wrote:

> Jeffrey Turner <[email protected]> wrote
>
>
>>>The minimal definition of evolution is "a change in the
>>>heritable characteristices of a population over time."
>>>This does not require speciation. The change can be due
>>>to random genetic drift or natural selection or possibly
>>>other mechanisms. All scientists agree that it's possible
>>>to have evolution without speciation and without
>>>adaptation.
>>
>>Yeah, well. Given that loose a definition, sure it is.
>>Drift is ubiquitous. It just isn't very important in the
>>real world. Be glad you've got college boards of trustees
>>to pay your salaries.
>
>
> <snip>
>
>>>We shouldn't have to be defending random genetic drift
>>>against those who would dismiss it as a mechanism of
>>>evolution.
>>
>>Electrical circuits all have noise, too. In understanding
>>the mechanisms we generally time average measurements so
>>that noise doesn't get in the way of our understanding of
>>fundamental principles. Go study your genetic noise, I
>>won't interrupt your navel gazing again.
>>
>>
>>>Those sorts of things should be moved to talk.origins or
>>>some other newsgroup that's intended for people who don't
>>>know much about evolution.
>>
>>Zounds, thou hast cut me to the quick. I know enough about
>>evolution to realize that genetic drift doesn't explain
>>anything about it. Drift happens, yawn.
>
>
> Jeffrey, you're a breath of fresh air here in SBE.
>
> Here's another approach you might take in opposition to
> this nonsensical supposition that drift represents a form
> of evolution distinct from natural selection. As you know
> they claim that genetic drift is the result of random
> causation. And if we follow the logic of this argument
> that would mean that natural selection must be the result
> of nonrandom causation. Ask them to substantiate this. You
> won't get a response.

Unfortunately, it's not worth the effort. Science was
supposed to explain the way things are. Random genetic drift
shrugs its shoulders and says there is no explanation. But
what they are studying is insignificant. All the major
changes in evolution happened much too quickly for random
drift to be a satisfactory explanation. So drift is an
excellent explanation of why some people have blue eyes and
some green. And whoever drew up the definition of evolution
did it in such a way that that is within the study of
evolution. Thus science dies not with a bang but with a
whimper. Ho hum.

--Jeff

--
A man, a plan, a cat, a canal - Panama!

Ho, ho, ho, hee, hee, hee and a couple of ha, ha, has;
That's how we pass the day away, in the merry old land of
Oz.
 
r norman <rsn_@_comcast.net> wrote:-

>>>>JE:- Typically BOH replaces testable reality with just
>>>>his "expectations" of it. In simple terms such an absurd
>>>>event is termed: propaganda. The master was Stalin. He
>>>>turned reality into his "expectations" and millions
>>>>perished.

>>>BOH:- John, please read up on simple probability theory
>>>before you write another post ridiculing expectations.
>>>They have a precise mathematical definition, which you
>>>seem to be totally unaware of. As for testability, you
>>>might like to read up about a wonderful subject called
>>>"statistics". It's used to fit models to data, which is
>>>an esential component of testing hypotheses. One of the
>>>things you can do is to estimate expected values (e.g. of
>>>allele frequency changes).

>>>JE:- Once again BOH asks us to replace a testable reality
>>>with just his expectations of it.

>> <snip>

>> BOH:- This is just a sematic game.

>> JE:- No, it is your "game" of misrepresentation.

> BOH:- No, your statement confuses two uses of the word
> "expectation". I was using the word in a technical sense,
> which I then tried to explain. Your senetence seemed to
> use the non-technical sense of the word, which is a
> different matter. Your sentence implies that in replacing
> the real world with something else, I am not making any
> allowance for the fact that the real world may be
> different. This is not the case, and the implication in
> your sentence comes purely from the different uses of a
> word. In that sense it is purely a matter of playing with
> semantics.

JE:- If you are not just "replacing the real world with
something else" using the "technical sense" of the word
"expectation", then what is the "real world" representation
of fitness that you are estimating?

><snip>

>> BOH:- The expectation is of a random variable, not of
>> reality.

>> JE:- Which makes your attempt to replace a testable
>> Darwinian reality with such an estimate, even more
>> absurd.

>BOH The only "Darwinian reality" is the real world. In our
>theories we have to replace the real world with something
>simpler, otherwise we cannot understand it.

JE:- Incorrect. Our theories of the world constitute all
of our _concepts_ of it. We are not granted any god given
concepts about our world, just percepts. We have to make
up all our _concepts_ for ourselves. We do this by
creating _contestable_ theories of _nature_ so that we can
eliminate the poorer ones. This is completed via the
refutation of absolute assumptions contained within each
concept we create.

RN:- Bob, John Edser simply refuses to accept as scientific
anything that involves probability or probabilistic
phenomena or processes.

JE:- Incorrect. All statistics can do is establish a
correlation. Such correlations can indicate if any observed
pattern is significant or not significant, i.e. is just a
non significant random OR a significant, non random pattern.
In the sciences the inductive (creative) mind must now
explain any non random correlation via the invention of
contesting theories of cause and effect. Always, _four_
possibilities exist. If we assume we have observed two non
random patterns A and B such that these patterns are also
correlated in a non random way then:

1) The process that caused pattern A caused the process that
caused pattern B

2) The process that caused pattern B caused the process that
caused pattern A

3) A 3rd unknown process c, caused both of the other
processes.

4) None of the processes that caused pattern A or pattern B
are associated in a causative way.

Only one of the 4 above, must remain.

RN:- I don't know what he does with quantum theory or
analysis of membrane noise or diffusion or anything related.
It doesn't do any good arguing.

JE:- The above form valid _parts_ of contestable theories of
nature, nothing more and nothing less.

RN:- If you think this is strange, you should read his stuff
on set theory!

JE:- All I suggested was:

(5) All Darwinian parents have a finite total fitness within
one population. This the total count of all fertile
forms reproduced by each parent. Each parental total
fitness count is the absolute fitness of that parent. It
remains fixed.

(6) Each parent within each population compares their total
fitness count with every other via just a default
process, i.e. a process that requires no cognition and
no energy. This default comparison must compare at least
two parental absolute fitness counts,
i.e. it is the relative difference between at least two
total parental fitnesses.

(3) The parent/s with the largest absolute fitness
is/are automatically selected.

4) A logic of why (3) is true MUST BE PROVIDED. As yet Neo
Darwinians do not think that automatic selection needs to
be explained because it is just, "obvious". The more
obvious something is the more obvious it is to any
competent epistemologist that a valid logic MUST be must
provided to understand
it.

5) To provide a logic for the natural selective event I have
described above, I simply proposed that set theory can be
validly be employed to illustrate
iu. In this simple illustration each parents total fitness
count, i.e. its absolute fitness now constitutes just a
simple _numerical_ total. These numerical totals are
totally intersected. The numerical total that is the
largest becomes a single set containing all the others
(in this case just one other) as sub sets.

Example: If A = total 7 and B = total 10 then when they are
totally intersected the following simple Venn Diagram is
produced. It shows the total intersection of two circles
representing set A and B which are just sets of numbers.

Venn Diagram:

---- ----
/ \/ \
/ /\ \
| | | |
| |7 | 3 |
| | | |
\ \/ / \ /\ / ---- ---- set A Set B

ALL of set A is now totally within set B so that only set B
actually exists to be selected, i.e. only one set actually
exists for nature to select so that nature has no choice but
to select set B. In so doing, nature is also selecting set A
but NOT, simultaneously. Sub set A is sub selected after set
B has firstly been selected, i.e. set A comes 2nd in natures
automatic natural selective game.

It makes no difference how many sets are intersected,
i.v. more than just two can also be intersected. If the two
sets have exactly the same totals, say 10, then:

---- ---- / \/ \ / /\ \
| | | |
| |10| 0 |
| | | |
\ \/ / \ /\ / ---- ---- set A Set B

In this instance both sets are selected simultaneously so
that both of them win.

The exact orders of natural selection are easily described
using this SIMPLE logical illustration. When n unequal sets
are totally intersected only 1 set is selected but n-1 sets
are _sub_ selected in an exact logical sequence.

Neither RN, AW, JM, BOH or GH who all contributed to
the thread:

"Intersecting Sets Of Fitness"

have contributed any valid criticism of the above which in
essence just compares two numerical totals. IMHO
intellectual honesty is seems to be missing re: this
discussion.

___________________________________________________
Is it valid or in valid to compare two or more sets of
numbers in the way described? I would appreciate an
honest ANSWER,
____________________________________________________

Respectfully,

John Edser Independent Researcher PO Box 266 Church Pt NSW
2105 Australia

[email protected]
 
I think that the comments below represent a gross
misunderstanding of genetic drift and a false distinction
between drift and selection based explanations of evolution.

in article [email protected], Jeffrey Turner
at [email protected] wrote on 3/25/04 8:49 AM:

> Science was supposed to explain the way things are. Random
> genetic drift shrugs its shoulders and says there is no
> explanation.

Drift is as much an explanation as selection. It is not
giving up, nor is it a mysterious black box. Drift is based
on the physical reality of sampling error. Sampling error is
as real and meaningful a source of change as say patterns of
predation. Indeed, the effects of sampling error are far
easier to characterize, and therefore drift-based models
make more precise quantitative evolutionary predictions than
selection-based explanations and are more testable. Indeed,
the precision of drift-based models makes them easier to
reject in empirical research than selection-based models.
This is evidenced by our increasingly powerful ability to
reject drift models when selection has had even a very small
influence over the evolution of DNA sequences. The general
premise that selection has driven evolution is entirely
untestable at the moment. You might be able to reject one
narrowly defined selection model in a given instance, but
selectionists tend to take this as evidence that some other
selection model must explain the data. To be rigorous, we
are relegated asking how much or little effect has selection
had on the evolution of a DNA sequence. The study of
evolutionary "forces" at the whole organism level is greatly
constrained in comparison, because we don't have a precise
model of drift effects at that level.

Note the following study to support my argument:

Experimental evolution yields hundreds of mutations in a
functional viral genome Bull JJ, Badgett MR, Rokyta D,
Molineux IJ JOURNAL OF MOLECULAR EVOLUTION 57 (3): 241-
248 SEP 2003

Jim Bull and colleagues established parallel (isolated)
lines of the T7 bacteriophage and kept each line at a very
small effective population size. They were able to sequence
entire genomes and document 404 mutational fixations in one
line and 299 in another line. As predicted by the drift-
based model, and contrary to any reasonable selection-based
model, many more of these base substitutions eroded phage
function by altering coding for amino acids than is
observed in populations maintained at larger effective
population sizes. Indeed, the authors used these results to
suggest that this simple procedure can be used to yield
attenuated viral strains for use as vaccines, as opposed to
the traditional method of passing viral populations through
non-human host species. The traditional method induces
selection on the virus to adapt to the novel host and
results in a "live" viral strain that can induce human
immune response without being virulent in humans. The
method of Bull et al. induces a more generic lack of
function by inducing strong drift.

Now, if you want to argue that small effective population
size results in unpredictable directions of selection, as
opposed to what we call drift, I would argue that such an
explanation is inconsistent with the primary result of the
viral study. This is because whatever direction selection
takes ought to improve function.

Drift is not a claim of effect without cause; rather, it is
a claim that sampling error is a generic result of sampling
processes, regardless of the non-systematic diversity of
mechanical causes affecting the outcome of each item
selected. As argued above, sampling error is a sufficiently
real phenomenon that its effects can be concretely
characterized. I would also argue that the process of
selection has an ontologically equivalent foundation as an
explanation. If, for example, you want to claim that
predation pressure caused the evolution of a species in some
way, then you are arguing that this was a net effect of lots
of individual events (e.g., predation events, escape events,
...). If the net effect was to make the prey species more
cryptic, it may still be that some unusually cryptic
individuals were preyed upon along the way. It may also be
that some less cryptic individuals were preyed upon for
reasons other than their degree of crypticity. It is the net
effect of all these roughly independent events that we call
selection, just as it is the net effect of the sampling
individual items that we call drift.

> But what they are studying is insignificant. All the major
> changes in evolution happened much too quickly for random
> drift to be a satisfactory explanation.

Now that is a quantitative claim. Is it just a gut feeling
of yours, or can you provide evidence to support your claim?
The Bull et al. study described above showed that evolution
occurred faster in the high-drift condition, including the
evolution of what they (plausibly, IMHO) call "speciation-
like" evolution.

Cheers,

Guy
 
On Fri, 26 Mar 2004 01:20:47 +0000 (UTC), "John Edser"
<[email protected]> wrote:

>r norman <rsn_@_comcast.net> wrote:-
>
<snip> John Edser's non-traditional style of quoting and non-
bottom posting makes snipping with context difficult.

>RN:- I don't know what he <John Edser> does with quantum
>theory or analysis of membrane noise or diffusion or
>anything related. It doesn't do any good arguing.
>
>JE:- The above form valid _parts_ of contestable theories
>of nature, nothing more and nothing less.
>

Since these all necessarily involve random processes, that
is they are probabilistic rather than deterministic
phenomena, then why not accept drift as a valid part of
evolution? Note, incidentally, that the word "drift" when
applied to a random process is usually the opposite of the
usage of the word "drift" applied to evolution.

I have snipped John's repeat of his unusual and unorthodox
(to say the least) ideas on fitness and how they might be
modeled using set theory.
 
John Edser wrote:
> r norman <rsn_@_comcast.net> wrote:-
>
>
>>>>>JE:- Typically BOH replaces testable reality with just
>>>>>his "expectations" of it. In simple terms such an
>>>>>absurd event is termed: propaganda. The master was
>>>>>Stalin. He turned reality into his "expectations" and
>>>>>millions perished.
>>>>
>
>>>>BOH:- John, please read up on simple probability theory
>>>>before you write another post ridiculing expectations.
>>>>They have a precise mathematical definition, which you
>>>>seem to be totally unaware of. As for testability, you
>>>>might like to read up about a wonderful subject called
>>>>"statistics". It's used to fit models to data, which is
>>>>an esential component of testing hypotheses. One of the
>>>>things you can do is to estimate expected values (e.g.
>>>>of allele frequency changes).
>>>
>
>>>>JE:- Once again BOH asks us to replace a testable
>>>>reality with just his expectations of it.
>>>
>
>
>>><snip>
>>
>
>>>BOH:- This is just a sematic game.
>>
>
>
>>>JE:- No, it is your "game" of misrepresentation.
>>
>
>>BOH:- No, your statement confuses two uses of the word
>>"expectation". I was using the word in a technical sense,
>>which I then tried to explain. Your senetence seemed to
>>use the non-technical sense of the word, which is a
>>different matter. Your sentence implies that in replacing
>>the real world with something else, I am not making any
>>allowance for the fact that the real world may be
>>different. This is not the case, and the implication in
>>your sentence comes purely from the different uses of a
>>word. In that sense it is purely a matter of playing with
>>semantics.
>
>
> JE:- If you are not just "replacing the real world with
> something else" using the "technical sense" of the word
> "expectation", then what is the "real world"
> representation of fitness that you are estimating?
>
My phrasing wasn't so clear - I am "replacing the real world
with something else", but that I can still make allowances
for it being different. My point was that when "taking
expectations", I'm not asking you to believe what I think
will happen, instead I'm integrating a random variable over
it's density function.

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
 
Guy Hoelzer wrote:
> I think that the comments below represent a gross
> misunderstanding of genetic drift and a false distinction
> between drift and selection based explanations of
> evolution.
>
> in article [email protected], Jeffrey
> Turner at [email protected] wrote on 3/25/04 8:49 AM:
>
>
>>Science was supposed to explain the way things are. Random
>>genetic drift shrugs its shoulders and says there is no
>>explanation.
>
>
> Drift is as much an explanation as selection. It is not
> giving up, nor is it a mysterious black box. Drift is
> based on the physical reality of sampling error. Sampling
> error is as real and meaningful a source of change as say
> patterns of predation. Indeed, the effects of sampling
> error are far easier to characterize, and therefore drift-
> based models make more precise quantitative evolutionary
> predictions than selection-based explanations and are more
> testable. Indeed, the precision of drift-based models
> makes them easier to reject in empirical research than selection-
> based models. This is evidenced by our increasingly
> powerful ability to reject drift models when selection has
> had even a very small influence over the evolution of DNA
> sequences. The general premise that selection has driven
> evolution is entirely untestable at the moment. You might
> be able to reject one narrowly defined selection model in
> a given instance, but selectionists tend to take this as
> evidence that some other selection model must explain the
> data. To be rigorous, we are relegated asking how much or
> little effect has selection had on the evolution of a DNA
> sequence. The study of evolutionary "forces" at the whole
> organism level is greatly constrained in comparison,
> because we don't have a precise model of drift effects at
> that level.
>
> Note the following study to support my argument:
>
> Experimental evolution yields hundreds of mutations in a
> functional viral genome Bull JJ, Badgett MR, Rokyta D,
> Molineux IJ JOURNAL OF MOLECULAR EVOLUTION 57 (3): 241-248
> SEP 2003
>
> Jim Bull and colleagues established parallel (isolated)
> lines of the T7 bacteriophage and kept each line at a very
> small effective population size.

Oh, evolution in a box - how clever.

> They were able to sequence entire genomes and document 404
> mutational fixations in one line and 299 in another line.
> As predicted by the drift-based model, and contrary to any
> reasonable selection-based model, many more of these base
> substitutions eroded phage function by altering coding for
> amino acids than is observed in populations maintained at
> larger effective population sizes.

How can you compare random genetic mutations in an
artificial environment to natural selection in the real
world? I'm not going to argue that genetic mutations
don't occur. That's the raw stuff that natural selection
acts upon.

> Indeed, the authors used these results to suggest that
> this simple procedure can be used to yield attenuated
> viral strains for use as vaccines,

Yes, humans will impose _selection_ on these bacteria to
achieve large populations of those that have advantageous
atributes. This is called "breeding" and humans have been
doing it to other species for thousands of years.

> as opposed to the traditional method of passing viral
> populations through non-human host species. The
> traditional method induces selection on the virus to adapt
> to the novel host and results in a "live" viral strain
> that can induce human immune response without being
> virulent in humans. The method of Bull et al. induces a
> more generic lack of function by inducing strong drift.
>
> Now, if you want to argue that small effective population
> size results in unpredictable directions of selection, as
> opposed to what we call drift, I would argue that such an
> explanation is inconsistent with the primary result of the
> viral study. This is because whatever direction selection
> takes ought to improve function.
>
> Drift is not a claim of effect without cause; rather, it
> is a claim that sampling error is a generic result of
> sampling processes, regardless of the non-systematic
> diversity of mechanical causes affecting the outcome of
> each item selected. As argued above, sampling error is a
> sufficiently real phenomenon that its effects can be
> concretely characterized. I would also argue that the
> process of selection has an ontologically equivalent
> foundation as an explanation. If, for example, you want to
> claim that predation pressure caused the evolution of a
> species in some way, then you are arguing that this was a
> net effect of lots of individual events (e.g., predation
> events, escape events, ...). If the net effect was to make
> the prey species more cryptic, it may still be that some
> unusually cryptic individuals were preyed upon along the
> way. It may also be that some less cryptic individuals
> were preyed upon for reasons other than their degree of
> crypticity. It is the net effect of all these roughly
> independent events that we call selection, just as it is
> the net effect of the sampling individual items that we
> call drift.
>
>>But what they are studying is insignificant. All the major
>>changes in evolution happened much too quickly for random
>>drift to be a satisfactory explanation.
>
> Now that is a quantitative claim. Is it just a gut feeling
> of yours, or can you provide evidence to support your
> claim? The Bull et al. study described above showed that
> evolution occurred faster in the high-drift condition,
> including the evolution of what they (plausibly, IMHO)
> call "speciation-like" evolution.

Breeding in controlled environments is hardly earth-
shattering. As far as supporting my claim, I used a random
genetic drift calculator I found on the net,
http://www.utm.edu/~rirwin/Drift.htm and, used the maximum
population size of 200 and a mutation frequency of 0.005
corresponding to a mutation in a single individual. In 50
trials, one mutation lasted for the full maximum duration of
200 generations and was at about 53% at the end having maxed
out under 66%. When Europeans started settling North America
in numbers the bison population was 60 million. To evolve a
random characteristic in a population that size to the point
where a new species would have been 100% different from the
progenitor would have taken more time than the existence of
the universe if you assume a reasonable length of a
generation and suitable scaling of the numbers for the
artificially small population of 200 used in the
mathematical exercise.

--Jeff

--
A man, a plan, a cat, a canal - Panama!

Ho, ho, ho, hee, hee, hee and a couple of ha, ha, has;
That's how we pass the day away, in the merry old land of
Oz.
 
"Guy Hoelzer" <[email protected]> wrote in message
news:[email protected]...
> I think that the comments below represent a gross
> misunderstanding of genetic drift and a false distinction
> between drift and selection based explanations of
> evolution.
>
> in article [email protected], Jeffrey
> Turner at [email protected] wrote on 3/25/04 8:49 AM:
>
> > Science was supposed to explain the way things are.
> > Random genetic
drift
> > shrugs its shoulders and says there is no explanation.
>
> Drift is as much an explanation as selection. It is not
> giving up, nor is it a mysterious black box. Drift is
> based on the physical reality of sampling error. Sampling
> error is as real and meaningful a source of
change
> as say patterns of predation. Indeed, the effects of
> sampling error are
far
> easier to characterize, and therefore drift-based models
> make more precise quantitative evolutionary predictions
> than selection-based explanations
and
> are more testable. Indeed, the precision of drift-based
> models makes them easier to reject in empirical research
> than selection-based models. This
is
> evidenced by our increasingly powerful ability to reject
> drift models when selection has had even a very small
> influence over the evolution of DNA sequences. The general
> premise that selection has driven evolution is entirely
> untestable at the moment. You might be able to reject one
narrowly
> defined selection model in a given instance, but
> selectionists tend to
take
> this as evidence that some other selection model must
> explain the data.
To
> be rigorous, we are relegated asking how much or little
> effect has
selection
> had on the evolution of a DNA sequence. The study of
> evolutionary
"forces"
> at the whole organism level is greatly constrained in
> comparison, because
we
> don't have a precise model of drift effects at that level.
>
<snip>

There is of course a basic-most patterning variability
(uncertainty) in Nature upon which the 'Natural Selection
part' of DNA-evolution has built error-correcting and even
sophisticated selectively mutation-rate controling genomes.

IMO this is a reality which is beautifully and convincingly
being drawn attention to by Lynn Helen Carporale in her
recent article in New Scientist, and, presumably, in her
book "Darwin in the Genome" - that I have not read.)

That drift - as both basic variability and sampling error -
might be easier to quantify than "selection", is hardly
anything controversial.

So instead, the fact that selection (of differently fit
genophenotypes) is AS HARD to quantify, as it is (unless one
is Edser), should be a wake-up call for people to try some
new theoretical tacks.

What I suggest is: to look at Evolution [both as _first
emergences_ of physical and biochemical patterns/patterning
tendencies and as preexisting (already emerged) patterns
that can be seen to perform Darwinian selection] in a
complementary new (?) explanatory
philosophical/terminological way;

This "way" involves tools for thought such as the
"opportunity type pressures" and "adversity type pressures";
and strategically categorized sum-over-histories - in this
case generalizable features of life-situations (or
environments) affecting individuals of to us (and to other
animals) phylogenetically significant populations.

This approach is more a strategy for recognizing new
explanatory patterns within already available fields of
view, than anything like a new theoretical recipe for
performing a never before tried experiment to confirm or
refute some predicted quantitative measurement (such as
Einstein's two relativity theories were).

P
 
On Sat, 27 Mar 2004 04:14:04 +0000 (UTC),
Jeffrey Turner <[email protected]> wrote:

[snip]

> Breeding in controlled environments is hardly earth-
> shattering. As far as supporting my claim, I used a random
> genetic drift calculator I found on the net,
> http://www.utm.edu/~rirwin/Drift.htm and, used the maximum
> population size of 200 and a mutation frequency of 0.005
> corresponding to a mutation in a single individual. In 50
> trials, one mutation lasted for the full maximum duration
> of 200 generations and was at about 53% at the end having
> maxed out under 66%. When Europeans started settling North
> America in numbers the bison population was 60 million. To
> evolve a random characteristic in a population that size
> to the point where a new species would have been 100%
> different from the progenitor would have taken more time
> than the existence of the universe if you assume a
> reasonable length of a generation and suitable scaling of
> the numbers for the artificially small population of 200
> used in the mathematical exercise.

If what you say is true then population geneticists and
evolutionary biologists must be really, really, stupid to
believe in random genetic drift. Is that what you think?

When your conclusion is completely at odds with the
scientific consensus, there are three main possibilities:

(1) you're on to something new and original that nobody
has ever thought of
(2) you've made a mistake or don't know enough about
the subject
(3) you're a kook.

Which category do you see yourself in?

Larry Moran
 
J:-
>snip<
Science was supposed to explain the way things are. Random
genetic drift shrugs its shoulders and says there is no
explanation. But what they are studying is insignificant.
All the major changes in evolution happened much too quickly
for random drift to be a satisfactory explanation. So drift
is an excellent explanation of why some people have blue
eyes and some green. And whoever drew up the definition of
evolution did it in such a way that that is within the study
of evolution. Thus science dies not with a bang but with a
whimper. Ho hum.

JE:- Science does not "die", it simply evolves like
everything else does that exists within a biological world.
Gene centric Neo Darwinists prefer to model the gene as a
valid unit of selection within nature when it has always
been known that selection can only operate on a genes
_phenotype_. All phenotypes, without exception, are organism
_fitness_ dependent. This is why the organism concept
remains the most basic concept within the science of
biology. The organism is the point of fitness linkage for
every genomic gene. "Evolution" via genetic drift attempts
to delete the Darwinian organism concept and substitute it
with "the gene", deleting all genetic epistasis (fitness
linkage) in the process. Such unlikely events are a part of
a valid modelling process that however, can and have been,
grossly misused.

It is logically true that evolution by natural selection
could work without any drift at all but evolution that only
employs random sampling error (genetic drift) cannot work.
Thus random genetic drift can only supply temporal variation
and _not_ evolution, within and not separate (contestable)
to, the Darwinian theory of evolution by natural selection
which remains the only _testable_ theory of evolution that
we have, even today. Scientific rigor has been deleted from
most Neo Darwinian arguments. It has been replaced by each
specialist just pushing their own department as somehow
independent to all the others. No generalist exists to
diagnose this gross error. Consequently, each department can
misuse its own models with impunity. We have powerful Neo
Darwinian engines, smarter car bodies with indestructible
windscreens but not a single Neo Darwinian car that is
drivable. Meanwhile, Darwin's model T remains perfectly
drivable but quietly parked in the garage gathering rust...

respectfully,

John Edser Independent Researcher

PO Box 266 Church Pt NSW 2105 Australia

[email protected]
 
>>>>>JE:- Typically BOH replaces testable reality with just
>>>>>his "expectations" of it. In simple terms such an
>>>>>absurd event is termed: propaganda. The master was
>>>>>Stalin. He turned reality into his "expectations" and
>>>>>millions perished.

>>>>BOH:- John, please read up on simple probability theory
>>>>before you write another post ridiculing expectations.
>>>>They have a precise mathematical definition, which you
>>>>seem to be totally unaware of. As for testability, you
>>>>might like to read up about a wonderful subject called
>>>>"statistics". It's used to fit models to data, which is
>>>>an esential component of testing hypotheses. One of the
>>>>things you can do is to estimate expected values (e.g.
>>>>of allele frequency changes).

>>>>JE:- Once again BOH asks us to replace a testable
>>>>reality with just his expectations of it.

>>><snip>

>>>BOH:- This is just a sematic game.

>>>JE:- No, it is your "game" of misrepresentation.

>>BOH:- No, your statement confuses two uses of the word
>>"expectation". I was using the word in a technical sense,
>>which I then tried to explain. Your senetence seemed to
>>use the non-technical sense of the word, which is a
>>different matter. Your sentence implies that in replacing
>>the real world with something else, I am not making any
>>allowance for the fact that the real world may be
>>different. This is not the case, and the implication in
>>your sentence comes purely from the different uses of a
>>word. In that sense it is purely a matter of playing with
>>semantics.

> JE:- If you are not just "replacing the real world with
> something else" using the "technical sense" of the word
> "expectation", then what is the "real world"
> representation of fitness that you are estimating?

BOH:- My phrasing wasn't so clear - I am "replacing the real
world with something else", but that I can still make
allowances for it being different.

JE:- Please note that you failed to answer the question that
was asked.

What is supposed to correct what? If you ARE "replacing the
real world with something else" is the aim to correct your
replacement of the real world against a measure of a real
world event or do you intend to correct the real world
against your replacement of it?

BOH:- My point was that when "taking expectations", I'm not
asking you to believe what I think will happen, instead I'm
integrating a random variable over it's density function.

Is "integrating a random variable over it's density
function" corrected against a testable reality or are you
just suggesting such an event constitutes reality? If it is
supposed to constitute a fitness reality please answer the
original question that I asked:-

If you are not just "replacing the real world with something
else" using the "technical sense" of the word "expectation",
then what is the "real world" representation of fitness that
you are only estimating?

Regards,

John Edser Independent Researcher

PO Box 266 Church Pt NSW 2105 Australia

[email protected]
 
> JT:- Breeding in controlled environments is hardly earth-
> shattering. As far as supporting my claim, I used a random
> genetic drift calculator I found on the net,
> http://www.utm.edu/~rirwin/Drift.htm and, used the maximum
> population size of 200 and a mutation frequency of 0.005
> corresponding to a mutation in a single individual. In 50
> trials, one mutation lasted for the full maximum duration
> of 200 generations and was at about 53% at the end having
> maxed out under 66%. When Europeans started settling North
> America in numbers the bison population was 60 million. To
> evolve a random characteristic in a population that size
> to the point where a new species would have been 100%
> different from the progenitor would have taken more time
> than the existence of the universe if you assume a
> reasonable length of a generation and suitable scaling of
> the numbers for the artificially small population of 200
> used in the mathematical exercise.

LM:- If what you say is true then population geneticists and
evolutionary biologists must be really, really, stupid to
believe in random genetic drift. Is that what you think?

When your conclusion is completely at odds with the
scientific consensus, there are three main possibilities:

(1) you're on to something new and original that nobody
has ever thought of
(2) you've made a mistake or don't know enough about
the subject
(3) you're a kook.

Which category do you see yourself in?

JE:- Lets be absolutely self consistent... Which category do
you see _yourself_ in?

There is no need to act in such a blatantly arrogant way.
This list is not just about specialists using their own
jargon to talk among themselves it is mostly about
specialists informing the general public, who quite rightly
have many VALID questions to ask.

All you need to do is carefully and SIMPLY explain your
position. Above all, explain how an independent observer can
test your view for themselves. Your arrogance and failure to
partake in a simple thought experiment that can challenge
the accepted dogma that random sampling error, alone, can be
validly assumed to cause evolution and not just temporal
variation, puts into question your integrity.

Best Wishes,

John Edser Independent Researcher

PO Box 266 Church Pt NSW 2105 Australia

[email protected]
 
>RN:- I don't know what he <John Edser> does with quantum
>theory or analysis of membrane
noise or
>diffusion or anything related. It doesn't do any good
>arguing.

>JE:- The above form valid _parts_ of contestable theories
>of nature, nothing more and nothing less.

RN:- Since these all necessarily involve random processes,
that is they are probabilistic rather than deterministic
phenomena, then why not accept drift as a valid part of
evolution?

JE:- That was my point: drift is just "a valid part of
evolution". ERGO: Drift is *NOT* evolution it is just one
*PART* of evolution.

A part is not a whole, that is why it called a part! The
question remains: exactly what PART of evolution does drift
constitute?

RN:- Note, incidentally, that the word "drift" when applied
to a random process is usually the opposite of the usage of
the word "drift" applied to evolution.

JE:- I see no difference. The term "drift", only refers to a
random _pattern_; no more and no less. Its cause can be
EITHER a random or a non random process, no exceptions.

RN:- I have snipped John's repeat of his unusual and
unorthodox (to say the least) ideas on fitness and how they
might be modeled using set theory.

JE:- You all but called me a "kook" re: my "unusual and
unorthodox (to say the least) ideas on fitness and how they
might be modelled using set theory". What exactly was
"unorthodox" about me comparing a minimum of two numerical
totals using a set intersection in the way that I described?
AFAICS, such an event is just a conventional and simple
secondary school exercise in set theory using Venn Diagrams.

_____________________________________________________
To start with, just leave out what the numbers represent.
Do you agree or disagree that two numerical totals can be
validly be compared using the set intersection I
illustrated ? I am forced to put this very bluntly. If you
continue to refuse to answer this question sbe reader's
will have every right to question your integrity. Please
REPLY to this question.
______________________________________________________

Respectfully,

John Edser Independent Researcher

PO Box 266 Church Pt NSW 2105 Australia

[email protected]
 
On Sat, 27 Mar 2004 04:14:02 +0000 (UTC), "Anon."
<[email protected]> wrote:

>John Edser wrote:
>> r norman <rsn_@_comcast.net> wrote:-
>>
>>
>>>>>>JE:- Typically BOH replaces testable reality with just
>>>>>>his "expectations" of it. In simple terms such an
>>>>>>absurd event is termed: propaganda. The master was
>>>>>>Stalin. He turned reality into his "expectations" and
>>>>>>millions perished.
>>>>>
>>
>>>>>BOH:- John, please read up on simple probability theory
>>>>>before you write another post ridiculing expectations.
>>>>>They have a precise mathematical definition, which you
>>>>>seem to be totally unaware of. As for testability, you
>>>>>might like to read up about a wonderful subject called
>>>>>"statistics". It's used to fit models to data, which is
>>>>>an esential component of testing hypotheses. One of the
>>>>>things you can do is to estimate expected values (e.g.
>>>>>of allele frequency changes).
>>>>
>>
>>>>>JE:- Once again BOH asks us to replace a testable
>>>>>reality with just his expectations of it.
>>>>
>>
>>
>>>><snip>
>>>
>>
>>>>BOH:- This is just a sematic game.
>>>
>>
>>
>>>>JE:- No, it is your "game" of misrepresentation.
>>>
>>
>>>BOH:- No, your statement confuses two uses of the word
>>>"expectation". I was using the word in a technical sense,
>>>which I then tried to explain. Your senetence seemed to
>>>use the non-technical sense of the word, which is a
>>>different matter. Your sentence implies that in replacing
>>>the real world with something else, I am not making any
>>>allowance for the fact that the real world may be
>>>different. This is not the case, and the implication in
>>>your sentence comes purely from the different uses of a
>>>word. In that sense it is purely a matter of playing with
>>>semantics.
>>
>>
>> JE:- If you are not just "replacing the real world with
>> something else" using the "technical sense" of the word
>> "expectation", then what is the "real world"
>> representation of fitness that you are estimating?
>>
>My phrasing wasn't so clear - I am "replacing the real
>world with something else", but that I can still make
>allowances for it being different. My point was that when
>"taking expectations", I'm not asking you to believe what I
>think will happen, instead I'm integrating a random
>variable over it's density function.
>
>Bob

Your phrasing was entirely clear to anyone who has the
slightest training in probability.

Also note that John Edser's style of responding obscures the
history of the thread , traditionally represented by the '>'
symbols at the start of the line. Some snipping of
content,which is entirely appropriate, has completely
removed my comments from the previous exchange, even though
my name appears prominently in the history lines at the
start of the posting.