Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

How do body parts evolve?

0 views
Skip to first unread message

Jack

unread,
Apr 12, 2007, 4:10:27 PM4/12/07
to
All of the calculations of evolutionary probability I have seen assume
inherantly that each mutation provides a benefit and is therefore
retained by the advantage it brings (natural selection). Is there any
robust and proven demonstration that body parts can indeed evolve or
elongate to perform or better perform a desired function?

Consider a tail-less monkey: A tail would be beneficial but how could
a tail evolve when the first mutation produces no advantage? I don't
mean a small advantage, or a negligable advantage but actually zero
(possibly even a disadvantage). Even with 100 co-incidental
cummulative mutations in the area of tail growth over 100 generations
(vanishingly unlikely) you end up with a small lump - all pain, no
gain.

It seems to me that each mutation would have to produce startling
results in order to have any effect on survival which would not be
buried in the noise of other factors (chance, climate).

Pip R. Lagenta

unread,
Apr 12, 2007, 5:06:08 PM4/12/07
to
On 12 Apr 2007 13:10:27 -0700, "Jack" <caw...@gmail.com> wrote:
[snip]

>Is there any
>robust and proven demonstration that body parts can indeed evolve or
>elongate to perform or better perform a desired function?
[snip]

Evolution is observed. It happens to populations, not individuals.
Evolution does not operate based upon desire nor "desired functions".

In a population, if individuals who have a more elongated body part
produce more offspring than those with a less elongated body part,
then elongated body parts will occur more and more throughout the
population over time. This happens regardless of whether or not the
elongated body part better performs a "desired" function. It is a
matter of producing more offspring. If being better able to perform a
particular function results in more offspring, then being better able
to perform that particular function will occur more and more
throughout the population over time. This happens regardless of
whether or not a particular body part is elongated.

Perhaps some background information would be of use:
<http://darwiniana.org/intro1.htm#Introduction>
<http://en.wikipedia.org/wiki/Evidence_of_evolution>
<http://evolution.berkeley.edu/>
<http://evolution.berkeley.edu/evosite/evo101/index.shtml>
<http://evolution.mbdojo.com/evolution-for-beginners.html>
<http://home.houston.rr.com/bybayouu/Tenets_of_evolution.html>
<http://plesiosaur.com/creationism/index.php>
<http://www.abarnett.demon.co.uk/atheism/evolution.html>
<http://www.abarnett.demon.co.uk/atheism/factandtheory.html>
<http://www.creationtheory.org/Introduction/Evidence.shtml>
<http://www.csharp.com/creationism.html>
<http://www.don-lindsay-archive.org/creation/evo_science.html>
<http://www.ebonmusings.org/evolution/whatevoisnt.html>
<http://www.evolutionhappens.net/>
<http://www.freethoughtdebater.com/FEvolutionCase.htm>
<http://www.gly.uga.edu/railsback/1122science2.html>
<http://www.gly.uga.edu/railsback/1122science3.html>
<http://www.gly.uga.edu/railsback/1122science7.html>
<http://www.gly.uga.edu/railsback/1122sciencedefns.html>
<http://www.mansfield.ohio-state.edu/~sabedon/biol1510.htm>
<http://www.nap.edu/html/creationism/evidence.html>
<http://www.rothamsted.bbsrc.ac.uk/notebook/courses/guide/>
<http://www.skeptictank.org/hs/factfaq.htm>
<http://www.talkorigins.org/faqs/comdesc/>
<http://www.talkorigins.org/faqs/evolution-research.html>
<http://www.talkorigins.org/origins/faqs-qa.html>
<http://www.tolweb.org/tree/>
<http://www.uwgb.edu/dutchs/PSEUDOSC/Top10MythsEvol.HTM>
HTH

>
--
內躬偕爻,虜,齯滌`偕爻,虜,齯滌`偕爻,虜,齯滌`偕爻,虜,齯滌`偕爻,
Pip R. Lagenta Pip R. Lagenta Pip R. Lagenta Pip R. Lagenta
�虜,齯滌`偕爻,虜,齯滌`偕爻,虜,齯滌`偕爻,虜,齯滌`偕爻,虜,齯滌

-- Pip R. Lagenta
President for Life
International Organization Of People Named Pip R. Lagenta
(If your name is Pip R. Lagenta, ask about our dues!)
<http://home.comcast.net/~galentripp/pip.html>
(For Email: I'm at home, not work.)

geo...@hotmail.com

unread,
Apr 12, 2007, 5:25:25 PM4/12/07
to

It might be possible that tails could evolve in very small animals -
tadpoles, for example, where a tiny waggling bump might just give more
of an advantage in moving, and that larger animals could evolve from
this - it seems likely to me (a completely novice on matters of
evolution) that this is how mammals have tails - presumably our fish
ancestors benefited from being able to waggle their rear fin slightly
more than other fish.

Or perhaps a tail can be exapted from something else which isn't a
tail but benefits from being long - the peacock's tail, for example,
is actually the tail coverts grossly enlarged, although with sexual
selection it's easy to see how this may arise.

jensp...@hotmail.com

unread,
Apr 12, 2007, 6:03:21 PM4/12/07
to
On Apr 12, 10:10 pm, "Jack" <cawo...@gmail.com> wrote:
> All of the calculations of evolutionary probability I have seen assume
> inherantly that each mutation provides a benefit and is therefore
> retained by the advantage it brings (natural selection).

Absolute correct.

> Consider a tail-less monkey: A tail would be beneficial but how could
> a tail evolve when the first mutation produces no advantage? I don't
> mean a small advantage, or a negligable advantage but actually zero
> (possibly even a disadvantage). Even with 100 co-incidental
> cummulative mutations in the area of tail growth over 100 generations
> (vanishingly unlikely) you end up with a small lump - all pain, no
> gain.

Well that's not really how genes work. A single change will often
bring about a major difference, as some genes control the develpment
of several major bodyparts.
E.g. when a persons arms are short, that same person will in almost
every case also have short legs. The length of both upper and lower
extremeties are comtrolled by the same genes.
Or consider the case, that some people are born with a (complete)
sixth finger. It's not that their great-great-great..(etc)....-
grandfather had just a smal lump of a sixth finger, and that they now
after countless generations finally have a fully formed sixth finger.

> It seems to me that each mutation would have to produce startling
> results in order to have any effect on survival which would not be
> buried in the noise of other factors (chance, climate).

You shold also consider the fact, that many bodyparts are originally
evolved for a different purpose, and then later adopted for another
function (often serving a double purpose during a transition periode).
E.g. the fins of a fish becomes leges, then becomes "arms" as the
animal starts to walk only on the hind legs, and the becomes the
wings of a bird.
Your own arms, originally used for locomotion is now the seat of a
grasping and manipulation device known as hands. It was the
coincidence that your forefathers used to swing from branches that
allowed this transistion.

Evolution always has to do with already exists. This very non-design
effect is seen all over in biology.

J.O.


Robert Carnegie

unread,
Apr 12, 2007, 6:17:12 PM4/12/07
to
Try this experiment:
Sit in a chair.
Lean forward, reach down and touch the floor.
Now lean back a little.
Reach towards the floor without leaning forwards.
Within around thirty seconds you will be able to touch the floor.
Congratulations! You have elongated your arm to perform the function
of touching the floor.

Perplexed in Peoria

unread,
Apr 12, 2007, 6:46:33 PM4/12/07
to

"Jack" <caw...@gmail.com> wrote in message news:1176408627.5...@b75g2000hsg.googlegroups.com...

> All of the calculations of evolutionary probability I have seen assume
> inherantly that each mutation provides a benefit and is therefore
> retained by the advantage it brings (natural selection).

Close. But it would be much more accurate to say that each mutation which
gets 'fixed' in the population provides a benefit. No one denies, of
course, that a lot of mutations are harmful, and therefore are eliminated
by selection.

A more subtle point is that quite a few mutations get fixed by 'drift',
rather than by selection. This may not matter too much for your argument,
which is based on whether advantages can accumulate, but it is very important
for the people who look at DNA sequences and decide what is related to what.

> Is there any
> robust and proven demonstration that body parts can indeed evolve or
> elongate to perform or better perform a desired function?

There is plenty of artificial selection (animal breeding) evidence that
body parts can be elongated or shortened by selection. One famous example
is dachshunds, a breed of dogs artificially selected to hunt badgers.

Evolution of new body parts or organs takes a lot more time, so I don't
have any examples.

> Consider a tail-less monkey: A tail would be beneficial but how could
> a tail evolve when the first mutation produces no advantage? I don't
> mean a small advantage, or a negligable advantage but actually zero
> (possibly even a disadvantage). Even with 100 co-incidental
> cummulative mutations in the area of tail growth over 100 generations
> (vanishingly unlikely) you end up with a small lump - all pain, no
> gain.

Actually, you have chosen a bad example here, since the story is that
tail-less monkeys are known to have evolved from ancestors that had tails.
That might mean that your tail-less monkey could evolve a new tail with
just a few mutations. ;-)

And if you change the question to how did tetrapod animals in general
evolve tails, it turns out that tails go a very long way back - back to
when we were still fish!

Not only that, but some animals without tails, including us, actually have
tails at an embryonic stage of development. So the question becomes whether
it is advantageous for an embryo to keep its tail into childhood and adult
life. It isn't just the adult form of the organism which evolves. It is
the entire life history that evolves.

This is not just a quibble. It really does matter. Just about every
example of new organs or new body parts passes through a stage in its
development when even small differences may matter - they may matter a lot!
So, if someone is going to set up as a critic of evolution, they need to
learn a lot of biology just to know whether an example is a good one.

> It seems to me that each mutation would have to produce startling
> results in order to have any effect on survival which would not be
> buried in the noise of other factors (chance, climate).

Yeah, it might seem that way. But it is just not so. Even small differences
in fitness shine through the noise. Over enough generations, they make a
difference. The key is the fact that species population sizes are fairly
large. It turns out that the dividing line between the things that make
enough difference to matter and the things that really are swamped by noise
is pretty close to one over the population size. If there are ten million
of a particular kind of beastie out there, then any difference which saves
a beastie's life once every ten million lifetimes will be noticed by selection
and will (over thousands of generations) become fixed in the population.

But if the population is only ten thousand individuals, then that beneficial
(at the one in ten million level) mutation probably will not be fixed by
selection. There are literally thousands of research papers dealing with
this fact - proving it or using it. Scientists are neither stupid, nor rare.
Objections generated by amateurs like yourself have been thought about by
many generations of people who know what they are doing.

Robert Carnegie

unread,
Apr 12, 2007, 6:58:54 PM4/12/07
to
Perplexed in Peoria wrote:
> Scientists are neither stupid, nor rare.
> Objections generated by amateurs like yourself have been thought about by
> many generations of people who know what they are doing.

This doesn't mean that it's stupid to ask questions. It depends on
the question. We see some really dumb ones here!

Nic

unread,
Apr 12, 2007, 7:35:01 PM4/12/07
to
On 12 Apr, 23:46, "Perplexed in Peoria" <jimmene...@sbcglobal.net>
wrote:
> "Jack" <cawo...@gmail.com> wrote in messagenews:1176408627.5...@b75g2000hsg.googlegroups.com...

He's chosen a good example if the exercise is to see how something
secondarily lost can be tertiarily regained. Maybe it can't, and
doesn't happen.

I'm thinking what happens if it becomes useful to have a limb in a
certain place when there once was one there before? It seems that a
replacement structure can arise, but it is unlikely to be a homologous
restoration.

The example being the bottle nosed whale. One was recently stranded
in the Thames, and witnesses were mistaking it for a shark because it
had a dorsal fin shaped exactly like one you would see on a shark.
That is exactly the evolutionary scenario Jack is talking about -
something appearing where there was nothing before, and where a
miniscule instance of such an organ would make no survival difference
at all.

Note how whales have failed to lose some vestigial bones over the same
timescale in which they have evolved a whole new stabilisation 'fin'
de novo.

I can only conjecture that is easier to evolve new sticky-out bits if
they are on the mid line (cf. camels' humps). With the obvious
exception of breasts. So perhaps if an animal with no tail needed to
come up with one, it might make it out of fat, of out of stuck-
together hair like rhino horn.

Dick C

unread,
Apr 12, 2007, 8:03:35 PM4/12/07
to
Jack wrote in talk.origins

> All of the calculations of evolutionary probability I have seen assume
> inherantly that each mutation provides a benefit and is therefore
> retained by the advantage it brings (natural selection). Is there any
> robust and proven demonstration that body parts can indeed evolve or
> elongate to perform or better perform a desired function?
>
> Consider a tail-less monkey: A tail would be beneficial but how could
> a tail evolve when the first mutation produces no advantage?

You are going the wrong way. Monkeys did not evolve tails. The predecessors
of Monkeys had tails, their predecessors also had tails. And so on, almost
all the way back to the earliest multicell beings. Since they lived in the
water a tail, even a small one, has great advantages. Steering and
stability to name a couple off the top of my head.
Therefor, the problem is not how a tailless monkey evolved a tail, rather
it is how did apes lose theirs? A tail to an animal like an ape is of
little use. Knuckle walking and our version of bipedalism do not require
tails for balance, or steering. And the tails can get in the way, so that
any mutation that reduces the size of a tail is advantageous. Nor would it
have to happen in a single jump, or even a few short generations.

snip misunderstanding of the process.
I hope the above helps your understanding. When you see something that you
don't understand, try to come at it from another direction, and try to
learn about the history behind it, as I pointed out in this case.

--
Dick #1349
"They that can give up essential liberty to obtain a little temporary
safety deserve neither liberty nor safety."
~Benjamin Franklin

Home Page: dickcr.iwarp.com
email: dic...@comcast.net

Dan Luke

unread,
Apr 12, 2007, 8:12:42 PM4/12/07
to

"Jack" wrote:

> It seems to me that each mutation would have to produce startling
> results in order to have any effect on survival which would not be
> buried in the noise of other factors (chance, climate).

Oh, boy, do YOU have a lot of reading to do.

--
Dan

(who's done most of it...well, a lot of it...um, well, pretty much, anyway)


Nic

unread,
Apr 12, 2007, 8:37:58 PM4/12/07
to
On 13 Apr, 01:03, Dick C <foo.dic...@comcast.net> wrote:
> Jack wrote in talk.origins
>
> > All of the calculations of evolutionary probability I have seen assume
> > inherantly that each mutation provides a benefit and is therefore
> > retained by the advantage it brings (natural selection). Is there any
> > robust and proven demonstration that body parts can indeed evolve or
> > elongate to perform or better perform a desired function?
>
> > Consider a tail-less monkey: A tail would be beneficial but how could
> > a tail evolve when the first mutation produces no advantage?
>
> You are going the wrong way. Monkeys did not evolve tails. The predecessors
> of Monkeys had tails, their predecessors also had tails. And so on, almost
> all the way back to the earliest multicell beings. Since they lived in the
> water a tail, even a small one, has great advantages. Steering and
> stability to name a couple off the top of my head.
> Therefor, the problem is not how a tailless monkey evolved a tail, rather
> it is how did apes lose theirs? A tail to an animal like an ape is of
> little use. Knuckle walking and our version of bipedalism do not require
> tails for balance, or steering. And the tails can get in the way, so that
> any mutation that reduces the size of a tail is advantageous. Nor would it
> have to happen in a single jump, or even a few short generations.

I agree with you that Jack should be asking about the losing of,
rather than the gaining of tails in apes. I don't know why tails have
been lost. A tail is useful if it is prehensile. Jacob Bronowski
suggested a tail might be a bad thing because it might be used as a
crutch. Presumably he was implying selective pressure to avoid losing
a leg would be reduced if we had tails. If so, that would be a
macroevolutionary argument - i.e. selection for long term advantages
overriding selection for short term ones. If so, that is something
that is yet to achieve widespread acceptance in this group.

Thurisaz the Einherjer

unread,
Apr 12, 2007, 10:30:13 PM4/12/07
to
Jack off:

> Is there any
> robust and proven demonstration that body parts can indeed evolve or
> elongate to perform or better perform a desired function?

Is there a school you actually saw from the inside, ever?

--
Romans 2:24 revised:
"For the name of God is blasphemed among the Gentiles through you
cretinists, as it is written on aig."

My personal judgment of monotheism: http://www.carcosa.de/nojebus

Lorentz

unread,
Apr 12, 2007, 11:52:08 PM4/12/07
to

>
> Consider a tail-less monkey: A tail would be beneficial but how could
> a tail evolve when the first mutation produces no advantage? I don't
> mean a small advantage, or a negligable advantage but actually zero
> (possibly even a disadvantage). Even with 100 co-incidental
> cummulative mutations in the area of tail growth over 100 generations
> (vanishingly unlikely) you end up with a small lump - all pain, no
> gain.
In the case of tails, you got it backwards. The tail evolved with
the first vertebrates. Look at fish and protochordates. Most of them
have tails, with an obvious use. The tail was useful for swimming in
fish. When vertebrates crawled on land, it lost its original function.
The tail may have been useless just after transition to the land. But
mutations on vestigial organs or hardly used organs are not very
harmful. Random mutation along with natural selection has modified the
tail so that it is either used for other purposes or greatly
reduced.There was never a time when the monkee had to go through a
"just a lump" phase of tail evolution. Like many animals, the
preprimate and pretree living ancestor of the monkee may have used the
tail for balance.

Almost all reptiles have tails, birds have little tails, etc. etc.
Some tails are problematic. For instance, how come we have a tail?
Regardless, the pretail lump model for monkee evolution is a straw
man.

Starting with monkees indicates that you haven't really
investigated the isue carefully before coming up with it. The fact
that all vertebrates seem to have a tail to some degree was brought up
by Charles Darwin himself in Origin of the Species. The origin of a
tail per se has to do with the very early evolution of vertebrates. If
you want to take your argument back there, its okay by me. I am
leaving.

Steven J.

unread,
Apr 13, 2007, 2:27:01 AM4/13/07
to
On Apr 12, 3:10 pm, "Jack" <cawo...@gmail.com> wrote:
> All of the calculations of evolutionary probability I have seen assume
> inherantly that each mutation provides a benefit and is therefore
> retained by the advantage it brings (natural selection). Is there any
> robust and proven demonstration that body parts can indeed evolve or
> elongate to perform or better perform a desired function?
>
> Consider a tail-less monkey: A tail would be beneficial but how could
> a tail evolve when the first mutation produces no advantage? I don't
> mean a small advantage, or a negligable advantage but actually zero
> (possibly even a disadvantage). Even with 100 co-incidental
> cummulative mutations in the area of tail growth over 100 generations
> (vanishingly unlikely) you end up with a small lump - all pain, no
> gain.
>
This may, perhaps, be an explanation of why apes and tailless monkeys
have not re-evolved tails (as Lorenz notes, monkeys started out with
tails, and in some lineages lost them). On the other hand, there is
evidence that chickens still have the genes for making teeth, which
can be activated if exposed to the correct hormones. Human babies are
occasionally born with a stubby tail with cartilige rather than bone;
it is conceivable that there are mutations that could trigger a bony
tail of significant length in tailless anthropoids.

>
> It seems to me that each mutation would have to produce startling
> results in order to have any effect on survival which would not be
> buried in the noise of other factors (chance, climate).
>
This was Mivart's argument, back in the 19th century. It's important
to note that, in a large population, even tiny advantages (e.g. a
particular trait increases one's odds of reproductive success by 1%)
eventually add up. It's also worth noting that the initial advantage
of a feature may be different from the advantage that is provided by a
more derived form of that feature: a bird's wing, after all, is
modified grasping limb, which was a modified walking limb, which was a
modified swimming fin. It would not take a huge modification of a
primitive chordate's body plan to produce some sort of extension that
could be at least slightly useful in propulsion or steering.

-- Steven J.

Robert Carnegie

unread,
Apr 13, 2007, 6:38:34 AM4/13/07
to

And maybe - if we're being teleological - it could fill in with flesh
and bones later.

Often, genes that aren't expressed are intact but are switched off by
another gene. A mutation can reverse that and restore the original
gene - which, however, when not used, isn't conserved. It just isn't
selected /against/.

And I think someone just said we /do/ grow a tail - when we are a
foetus - and then un-grow it before birth. One to check and then
remember to use in arguments.

Btw I tried my arm-growing trick today when something I dropped rolled
under a car, but the process has a natural limit - I didn't achieve
macroevolution ;-) I went round the other side of the car and my
property was now within my Lamarckian scope!! (And my shirts still
fit for sleeve length. I didn't think of that...)

Message has been deleted

jensp...@hotmail.com

unread,
Apr 13, 2007, 7:49:23 AM4/13/07
to
So, we now have a fair number of rather informative replies. The only
thing missing is a commment from the original poster, as to whether he
now feels enlightned.
Strikes me as odd neither to ask further, nor to thank for a
sufficient reply.

J.O.


Dick

unread,
Apr 13, 2007, 8:35:17 AM4/13/07
to

I am fascinated with the peacock's tail. Especially the idea a hen
would find the colorful tail more sexually arousing. It makes me then
wonder how the hen's interest in this colorful display evolved?

It has been suggested that the hindrance of the tail to the male
proves to the hen he is strong enough to survive the encumbrance. I
guess a fat male would prove he was a good food provider.

Message has been deleted

Nic

unread,
Apr 13, 2007, 10:16:45 PM4/13/07
to
On 13 Apr, 12:27, nmp <addr...@is.invalid> wrote:
> Op Fri, 13 Apr 2007 03:38:34 -0700, schreef Robert Carnegie:

>
> > And I think someone just said we /do/ grow a tail - when we are a foetus
> > - and then un-grow it before birth.
>
> I don't think we really ungrow our tails. It only stops growing while the
> rest of our body continues.

Maybe they are not tails at all. When you put up a bookshelf, you
don't put the brackets at the very ends. You put them some way in
from the ends. I don't know what the optimum ratio is, but I suspect
it's a bit like a railway sleeper upside down.

So when you stick legs on a tetrapod, you wouldn't put them at the
extreme corners. That could lead to undue back strain.

Therefore, if you don't want to locate the back legs right at the very
back, you are bound to end up with a bit of an overhang. Tails don't
have to be explained as limbs in their own right. Think about
dinosaurs - don't you think they may have had one or two useful
giblets cantelevered out back?

> Then of course, in some people, it doesn't stop:
>
> <http://www.dimaggio.org/Archive/tails_in_humans.htm>


Kermit

unread,
Apr 13, 2007, 11:03:55 PM4/13/07
to
On Apr 13, 2:22 pm, nmp <addr...@is.invalid> wrote:
> Op Fri, 13 Apr 2007 04:49:23 -0700, schreef jenspolsen:
> Happens all the time, it shouldn't be a big deal. Perhaps Jack feels
> overwhelmed by the large number of responses and the speed with which
> talk.origins moves in general.

I forget about the relative speed sometimes. I have been to other
newsgroups that cover interesting topics, and when I start to post to
a thread, I look at the dates, and realize that a dozen or so posts
cover *months.

Kermit

Jack

unread,
Apr 23, 2007, 2:52:06 PM4/23/07
to
Well, thanks to everyone who took time to write a sensible reply. I'm
sorry my example of tail-less monkeys was not historically valid
although some of you seem to have understood the crux of my questions
which is this:
Is the process (evolution) which generates new features of survival/
reproductive benefit smooth or jagged. I think this is referred to as
the question of micro- or macro-evolution.

I had not considered that a single mutation could produce such
startling results. Would it be correct to say that a single mutation
could produce a whole new nipple on a pig thus allowing a mother to
feed an extra piglet thus producing comparatively more offspring thus
increasing the chances that this "extra nipple mutation" is passed on
and eventually "fixed"?

But how does this apply to entirely new appendages and organs? In the
case of a pig the "data" for building a nipple is already there
however, this data itself could not have arrived in one go. I suppose
this approaches the "but you can't have half an eye" debate but I
really do fail to see how an organism without say, an anus, could
evolve one. Either you have a miraculous jump of some sort
(macroevolution) or you have a smooth progression which, as I've
described, seems improbable in the extreme because of the *zero*
benefit of each supposed intermediate stage thus rendering it
improbable that these (non)benefits accumulate. I really am talking
about *zero benefit* steps not just small benefit steps.

So which model is correct: Macro- or Micro- evolution?

Perplexed in Peoria

unread,
Apr 23, 2007, 3:21:15 PM4/23/07
to

"Jack" <caw...@gmail.com> wrote in message news:1177354326.2...@e65g2000hsc.googlegroups.com...

> Well, thanks to everyone who took time to write a sensible reply. I'm
> sorry my example of tail-less monkeys was not historically valid
> although some of you seem to have understood the crux of my questions
> which is this:
> Is the process (evolution) which generates new features of survival/
> reproductive benefit smooth or jagged. I think this is referred to as
> the question of micro- or macro-evolution.

You think wrong. Micro- and macro-evolution are not competing theories.
They are either (depending on who you ask) two different processes producing
two different kinds of results - or else they are one process viewed on
two different time scales.

I think that the right way to characterize the question you are getting at
is to ask whether evolution is gradual or 'saltational'. The big name
associated with the 'saltational' idea is Richard Goldschmidt. Orthodox
modern Darwinism tends to accept gradualism, though the issue is confused
a bit by PE (punctuated equilibrium) which asserts that even gradual change
can be fast enough so that it can't be distinguished from saltation in the
fossil record.

> I had not considered that a single mutation could produce such
> startling results. Would it be correct to say that a single mutation
> could produce a whole new nipple on a pig thus allowing a mother to
> feed an extra piglet thus producing comparatively more offspring thus
> increasing the chances that this "extra nipple mutation" is passed on
> and eventually "fixed"?

Yes, it probably could. But whether the extra (pair of, probably) nipples
are advantageous would depend upon litter size, and other things. A pig
who produces small litters would probably find the extra nipples to be
nothing but a nuisance.

> But how does this apply to entirely new appendages and organs? In the
> case of a pig the "data" for building a nipple is already there
> however, this data itself could not have arrived in one go. I suppose
> this approaches the "but you can't have half an eye" debate but I
> really do fail to see how an organism without say, an anus, could
> evolve one.

Once again, you have chosen a bad example. We have had flow-thru digestive
tracks far longer than we have had legs, nipples, and even brains.

OlOlOl01

unread,
Apr 23, 2007, 4:01:24 PM4/23/07
to

"Jack" <caw...@gmail.com> wrote in message
news:1177354326.2...@e65g2000hsc.googlegroups.com...
Maybe neither. The biochemical processes involved possess
labyrinthine complexities and the science is yet infant and dimly
grasped.

I play around with Genetic Alhgorithms (a lot easier to grasp,
but using the same principles as biosystems) and might point the
following out (as patterned metaphors of the biochemistry):

1) speciation, in the form of building heritable, predominant patterns
(or elite chromosomes) occurs principally through crossover (biologically=
breeding or part chromosomic replication) and once in place _in a
stable ecology_ rapidly assumes strong dominance

2) to 'shift' a dominant pattern out of a stable population takes one
hell
of an effort in terms of mutation in that mutation has to be either:
a) extensive enough to disrupt the species stabilty and make new
groupings (and even a 5% mutation in every generation is often insufficient)
OR
b) sufficiently specific and powerful as to provide immediate superiority
beyond
the extant dominant chromosomes
Either of these could provide your miraculous jump...

3) the more replications of a given schema (=short pattern or 'gene')
that exist in the chromosomes of a population then the more likely
that chromosomic feature will occur (repeatedly) in subsequent
generations (this in part answers your number of nipples question:
nipples, as a biologicaly useful feature, (= a repeating
gene subset) will increase _in the given ecology_ until they reach
a useful maximum. They will then increase no more (and the optimum
wll vary synergetically as the ecology varies).
(as a corollory of this, consider the repeating pattern TTAGGG /number
of
telomeres on the tail end of a chromosome and 'useful' lifespan. I
speculate that repeats, and possibly palindromic repeats, in sequences are
markers of 'number' and temporality in 'number of repeats' (=duration))

4) it's interesting to run various sets of algorithms alongside each other
as an 'ecology' and to observe their ultimate stable forms (which like most
systems have various stable similar yet non-equal 'quantum' levels).
If you consider an individual organ of a body of any individual macro
'species' as itself being a species existing in the
ecology of that body (with certain genes in certain chromosomes
turned 'on'), and then look at the 'whole' of the planet/species/
organs/suborgans/chemical compounds etc. as a recursive,
synergystic set, it gives you a whole new view of things...

5) in Genetic Algorithms, you can play around with haploid, diploid,
tripolid,
indeed 'nploid' chromosomes as well as binary, quatenary, denary and even
grammatical encoding (the latter area is remarkably fruitful in terms of
innovation and relatively simple to handle in that language comes quite
naturally to human beings).
Biogeneticists should of course be playing around with diploid quatenary
sets of multiple chromosomes - but I, as a mathematician/engineer prefer
to mess about with the other stuff.

I heartily recommend study of these mathematical models since I feel it
can give insight to microbiology (and many other things evolutionary).

As an example, I have used the notion I express in 1) above to model
the behaviour of a rat chasing cheese in a maze.
A simple, haploid, six chromosome, 100 locus population
will usually track down the cheese in a 20 by 20 maze (with NSEW
possible movements and random start points for rat/cheese) with ten
or so generations (and you can very simply oprtimize the algorithm
such that it performs far better than that).

One then takes this beyond, lets the rat get part way to the
cheese & then randomly changes the location of the latter
(and/or the maze obstacles) whilst leaving the rat where it is &
letting the problem run on.
Ratty gets very ratty and very confused - indeed his 'genetic
learning' has been entirely disrupted by varying the ecology more
rapidly than he can adapt ( a fate that awaits us all if global
warning really gets hold)...

Now, this might seem a trivial problem - but when one
writes the following rough metaphor:
maze - building
maze obstacles - antibiotics
cheese - patients biosystems
rat - MRSA
rat learning - MRSA breeding

Then it takes on a whole new significance...

"A stable ecology in a hospital building provides ideal conditions for
a bacterial or viral disease to evolve and overcome the antibiotics in use
and attack the patients biosystems."

The beauty of metaphor.

John Harshman

unread,
Apr 23, 2007, 5:13:12 PM4/23/07
to
Perplexed in Peoria wrote:

And our ancestors had single-opening digestive tracts long before that.
Flow-through digestion has apparently been lost once too, in flatworms
(not counting all the parasites that have lost their digestive tracts
altogether).

As for the failure of imagination, imagine this: a polyp-like organism
that ingests and excretes through a single opening develops a somewhat
elongated opening, the two sides of which gradually specialize, one in,
out out. Now suppose that a constriction gradually evolves in the
middle, which partially separates the two functions, and that this
constriction gradually evolves into a partition, so there are two holes
where there was once one. Viola: a mouth and an anus, gradually evolved.
Now I don't know if there are any living intermediates, though I
wouldn't be surprised. But there are certainly analogous sorts of
intermediates in various other organs, including hearts, eyes, lungs,
and so on.

The general answer is that new features evolve from old features,
generally by successive slight modifications, and most features can
begin as little bumps, pockets, thickenings, or other simple shapes, not
the fancy organs we see at the end of the process.

Bill Morse

unread,
Apr 23, 2007, 9:18:23 PM4/23/07
to
John Harshman wrote:

Exactly. You have explained an anus, and the rest of the organs are easier
to understand. What use is a little bit of a kidney? Well if it helps
remove wastes more quickly than simple diffusion, it allows a higher
metabolic rate. It doesn't at first have to be either elaborate or
efficient - it just has to be better than no kidney. The heart can start
out (to use your term) as a thickening that contracts to speed up
circulation. The lungs can start as a pocket that increases surface area
and thus increases gas exchange.

As to micro vs. macro, be aware that as an elaboration to John's point about
new features evolving from old features, one of the very common genetic
processes by which this occurs is duplication of a set of genes. This
leaves the old genes to perform their original function while making a new
set that can change to perform a new function - but they already have a lot
of functionality, so the change can be relatively rapid, and can appear as
a "saltation".
--
Yours, Bill Morse

username_...@yahoo.com.au

unread,
Apr 24, 2007, 4:12:34 AM4/24/07
to
On Apr 13, 6:10 am, "Jack" <cawo...@gmail.com> wrote:
>
> It seems to me that each mutation would have to produce startling
> results in order to have any effect on survival which would not be
> buried in the noise of other factors (chance, climate).

Body parts actually evolve quite easily, especially limbs. Whales
lost their hind limbs, for instance, although that may have been a
fluke.

Klaus

unread,
Apr 24, 2007, 8:03:10 AM4/24/07
to
Jack wrote:
> All of the calculations of evolutionary probability I have seen assume
> inherantly that each mutation provides a benefit and is therefore
> retained by the advantage it brings (natural selection). Is there any
> robust and proven demonstration that body parts can indeed evolve or
> elongate to perform or better perform a desired function?
>
> Consider a tail-less monkey: A tail would be beneficial but how could
> a tail evolve when the first mutation produces no advantage? I don't
> mean a small advantage, or a negligable advantage but actually zero
> (possibly even a disadvantage). Even with 100 co-incidental
> cummulative mutations in the area of tail growth over 100 generations
> (vanishingly unlikely) you end up with a small lump - all pain, no
> gain.
>
> It seems to me that each mutation would have to produce startling
> results in order to have any effect on survival which would not be
> buried in the noise of other factors (chance, climate).
>

Monkeys started out with tails.
Klaus

OlOlOl01

unread,
Apr 24, 2007, 2:42:47 PM4/24/07
to

"Jack" <caw...@gmail.com> wrote in message
news:1177354326.2...@e65g2000hsc.googlegroups.com...

Jack

unread,
Apr 27, 2007, 6:40:41 AM4/27/07
to
On Apr 23, 11:13 pm, John Harshman <jharshman.diespam...@pacbell.net>

wrote:
> Perplexed in Peoria wrote:
> The general answer is that new features evolve from old features,
> generally by successive slight modifications, and most features can
> begin as little bumps, pockets, thickenings, or other simple shapes, not
> the fancy organs we see at the end of the process.- Hide quoted text -
But explaining that mutation occurs is only half the picture. Unless
you can justify it's transmission you don't get evolution. Thus, back
to my original question: how do these zero-benefit bumps accumulate?

I guess I am questioning the assumption that, frankly states, "an X is
a survival advantage, so a small X is a slight survival advantage".

Jack

unread,
Apr 27, 2007, 6:45:26 AM4/27/07
to
On Apr 24, 3:18 am, Bill Morse <wdNOSPAmo...@verizonOSPAM.net> wrote:
> John Harshman wrote:
> > Perplexed in Peoria wrote:
>
> >> "Jack" <cawo...@gmail.com> wrote in message
Indeed but it has to be a kidney and a "little kidney" is still a
kidney. I'm talking about the steps leading up to the point at which
it can function as a kidney. Lump of tissue does not qualify and falls
into the zero-benefit mutation category which should not proceed into
later generations.
Evolving organs/organisms as explained on this basis face a barrier
with say 7000 steps it has to cross before it get's even a basic
kidney. Now if each step is random, non-directional and potentially
fatal you are looking at some serious odds of failure. And I mean like
winning the lottery a trillion times kind of odds of failure.

Bob D

unread,
Apr 27, 2007, 7:01:00 AM4/27/07
to

Whales gained a flattened tail. That is the fluke!


Bob D

unread,
Apr 27, 2007, 7:32:16 AM4/27/07
to
> winning the lottery a trillion times kind of odds of failure.- Hide quoted text -
>
> - Show quoted text -

I think you are missing the point, Jack. There are many examples of
species alive today which have much 'simpler' excretory systems than a
fully fledged 'kidney'.

My memory in this area is hazy, but comparative anatomy, plus
embryological studies, give us a very good idea of how kidneys have
evolved.

I like this quote (from http://users.rcn.com/jkimball.ma.ultranet/BiologyPages/V/VertebrateKidneys.html):

"We like to think of ourselves as highly advanced. Why don't we have
kidneys as efficient as those of the reptiles and birds? It is the
luck of our inheritance. The line of vertebrate evolution that
produced the mammals split off before the evolution of the diapsids
whose ability to convert nitrogenous wastes into uric acid was passed
on to all their descendants, including the lizards, snakes, and birds.
"


John Harshman

unread,
Apr 27, 2007, 9:40:18 AM4/27/07
to
Jack wrote:

> On Apr 23, 11:13 pm, John Harshman <jharshman.diespam...@pacbell.net>
> wrote:
>
>>Perplexed in Peoria wrote:
>>The general answer is that new features evolve from old features,
>>generally by successive slight modifications, and most features can
>>begin as little bumps, pockets, thickenings, or other simple shapes, not
>>the fancy organs we see at the end of the process.- Hide quoted text -
>
> But explaining that mutation occurs is only half the picture. Unless
> you can justify it's transmission you don't get evolution. Thus, back
> to my original question: how do these zero-benefit bumps accumulate?

Who said they were zero-benefit bumps? In fact the bumps in question
must have some benefit or they don't accumulate.

> I guess I am questioning the assumption that, frankly states, "an X is
> a survival advantage, so a small X is a slight survival advantage".

Feel free to question anything. But why shouldn't a small X be a slight
survival advantage? For some organs, this has been worked out in detail,
e.g. Nilsson, D., and S. Pelger. 1994. A pessimistic estimate of the
time required for an eye to evolve. Proc. R. Soc. Lond. B 256:53-58. For
others it just seems obvious. A lung, for example, is merely an increase
in surface area and vascularization of some bit of pharynx. A little
indentation or fold in the pharynx, and a little increase in the number
of adjacent capillaries, would increase oxygen uptake of an organism
that had the habit of swallowing air. Just the habit alone, without any
increased area, would be useful to a fish that needed extra oxygen. So
there's an advantage from the start. And so on. Most "new" organs are of
this sort: merely increased surface area or some other simple
characteristic, achieved in simple ways.

Greg Guarino

unread,
Apr 27, 2007, 10:51:25 AM4/27/07
to
On Mon, 23 Apr 2007 21:13:12 GMT, John Harshman
<jharshman....@pacbell.net> wrote:

>Viola: a mouth and an anus

Make up your own musician joke.

Greg Guarino

Noelie S. Alito

unread,
Apr 27, 2007, 5:35:12 PM4/27/07
to

For the Journal-impaired among us, I offer an online illustration
of the Nilsson&Pelger work:


<http://www.don-lindsay-archive.org/creation/eye_stages.html>


Me like pictures!


Noelie
--
<myfirstname>@<capital_of_texas>.rr.com
"Rhyming with 'goalie' for over 46 years."

Greg G.

unread,
Apr 27, 2007, 5:48:46 PM4/27/07
to
On Apr 27, 10:51 am, Greg Guarino <g...@risky-biz.com> wrote:
> On Mon, 23 Apr 2007 21:13:12 GMT, John Harshman
>
> <jharshman.diespam...@pacbell.net> wrote:
> >Viola: a mouth and an anus
>
> Make up your own musician joke.

OK.
Q: What's the difference between a band and a bull?
A: The bull has the horns in front and the asshole in the back.
>
> Greg Guarino

--
Greg G.

If we had saxophone
Big baritones, cleaning up the muddy breaks
If we had saxophone
I could get some recognition from
that Mobile Alabama DJ
--Jimmy Buffett


Jack

unread,
Apr 30, 2007, 5:10:21 PM4/30/07
to
Perhaps you could share this good idea because for me the problem
remains: A kidney, or any other feature or organ is only beneficial in
the nth stage of "evolution" where n is a large number and steps 0
through n bring zero-benefit and are potentially fatal. You just don't
get single-cell kidneys or organs which are just like neighbouring
cells but for 1 or 2 minor adjustments which could well have occurred
by chance. The teleology seems undeniable.

Bill Morse

unread,
Apr 30, 2007, 9:55:35 PM4/30/07
to
Jack wrote:

> On Apr 27, 1:32 pm, Bob D <ju.d...@btopenworld.com> wrote:
>> On 27 Apr, 11:45, Jack <cawo...@gmail.com> wrote:

(snip)


>> > > Exactly. You have explained an anus, and the rest of the organs are
>> > > easier to understand. What use is a little bit of a kidney? Well if
>> > > it helps remove wastes more quickly than simple diffusion, it allows
>> > > a higher metabolic rate. It doesn't at first have to be either
>> > > elaborate or efficient - it just has to be better than no kidney.
>>
>> > Indeed but it has to be a kidney and a "little kidney" is still a
>> > kidney. I'm talking about the steps leading up to the point at which
>> > it can function as a kidney. Lump of tissue does not qualify and falls
>> > into the zero-benefit mutation category which should not proceed into
>> > later generations.
>> > Evolving organs/organisms as explained on this basis face a barrier
>> > with say 7000 steps it has to cross before it get's even a basic
>> > kidney. Now if each step is random, non-directional and potentially
>> > fatal you are looking at some serious odds of failure. And I mean like
>> > winning the lottery a trillion times kind of odds of failure.

>> I think you are missing the point, Jack. There are many examples of


>> species alive today which have much 'simpler' excretory systems than a
>> fully fledged 'kidney'.
>>
>> My memory in this area is hazy, but comparative anatomy, plus
>> embryological studies, give us a very good idea of how kidneys have
>> evolved.

> Perhaps you could share this good idea because for me the problem
> remains: A kidney, or any other feature or organ is only beneficial in
> the nth stage of "evolution" where n is a large number and steps 0
> through n bring zero-benefit and are potentially fatal. You just don't
> get single-cell kidneys or organs which are just like neighbouring
> cells but for 1 or 2 minor adjustments which could well have occurred
> by chance. The teleology seems undeniable.

Let us start with something like a ribbon worm (Nemertea - and note that I
am just using this as an example - someone like John Harshman would have a
better idea of the actual evolutionary history). Ribbon worms have a
circulatory system but no kidneys (as I understand it). Nevertheless, their
cells have to excrete waste products. Now let's suppose that a few of these
cells randomly become slightly specialized to remove wastes from the blood
vessels and excrete them, initially perhaps to the central body cavity that
leads to the anus. This is beneficial to the organism at the very get-go,
as it lowers the general concentration of waste products in the blood and
allows other cells to spend less energy rejecting wastes. It is non-fatal,
as the specialized cells are initially only experiencing slightly higher
concentrations of wastes. We have exactly what you say can't occur:
single-cell kidneys with minor adjustments from neighboring cells but which
still bring a net benefit at step 1.

--
Yours, Bill Morse

Martin Hutton

unread,
Apr 30, 2007, 11:20:52 PM4/30/07
to

Don't forget single celled organisms can maintain osmotic
balance and get rid of wastes. The basic biochemistry is
available in all cells and are required functions.

Your ideas that a feature is only important in the nth stage
and not (and possibly harmful) in the <nth stage and that
biochemistry cannot vary from neighbouring cells are false.

"Important" is not the word to use...try "useful" instead.

If a feature is harmful, then it will be selected out.
Neutral drift only fixes a few non-useful and non-harmful
features (hence neutral). If a feature is useful (even
only if it's a little teensy weensy tiny bit useful) it
will be selected for.

In a colony of single celled organisms the biochemistry of
cells on the edges does differ from those in the centre and
those in intermediate positions.

You might want to read an old POTM by R. Norman about
organ evolution:
(http://www.talkorigins.org/origins/postmonth/may06.html)
I'm sure that there's another POTM that's about kidneys
and their role in maintaining osmotic balance and getting
rid of nitrogenous waste...but I can't find it. I *think*
it was by Ron Okimoto...but I'm sure a T.O. regular will
correct me.

--
Martin Hutton

Jack

unread,
May 1, 2007, 11:06:55 AM5/1/07
to
I have the feeling that when you guys talk about a "small change"
we're still talking about hundreds if not thousands of what I call
"stages". I think it would help to clarify what I mean by "stage": I'm
talking about a single genetic mutation. If, as I understand from your
post, a single mutation is _generally_ able to produce dramatic
effects then we have a good case for macro-evolution: evolution
proceeding in jumps, seemingly unlikely but counter-balanced by sheer
numbers and time scale. It is my understanding however that only micro-
evolution (basically adaptation) is well supported by empirical
evidence and that the time-scales and numbers required for things like
chimps to become humans are ludicrously large and do not align with
the timescales from the fossil record by several orders of magnitude.

Bob D

unread,
May 1, 2007, 1:40:20 PM5/1/07
to
> the timescales from the fossil record by several orders of magnitude.- Hide quoted text -

>
> - Show quoted text -

So 5 million years is, what, too long? Too short? 500,000 generations
is too few for what kind of changes (assuming 10 years to sexual
maturity) required from a common ancestor with, say, a chimp?

According to creationists, 6,000 years is enough just to produce the
minor differences we see now between different populations of humans
globally (600 generations, to stick to the same measure. Fewer if you
allow for an increasingly delayed sexual maturity in humans). Multiply
that by almost 1,000. What differences might we then see? Especially
if you factor in a reproductive barrier, and intense selection
pressure, not yet experienced by extant groups of humans.

Macro vs micro evolution. How many micros, in your estimation, need to
be added together to make a macro?

As Chairman Mao was quoted as saying (and I'm probably misquoting,
though not in essence) 'Even the longest journey starts with a single
step'. Macro evolution is just loads of micros.

Jack

unread,
May 1, 2007, 3:20:26 PM5/1/07
to
Apparently it is (too short that is) and the math has been done. In
fact if you assume a 3% difference between the chimp and human genomes
(both not fully resolved to date) corresponding to 90 million base
pairs you start to see the challenge. Looking only at the active part
of the genome this reduces to by a further factor of 1%, or aprox. 1
million base pairs. Looks like we need a directional mutation every 2
generations. However, looking at the maximum measured rates of
mutation in cell division (1 in 10) or in the more robust gametes (1
in 10'000 - 100'000) the challenge becomes almost insurmountable. Now
you have to start assuming these mutations accumulate in a certain
direction (i.e. are survival-relevant), do not get lost or reversed by
later mutations and you start to see the miraculousness of us evolving
from chimps. Disclaimer: I'm not trying to cast doubt on evolution
occuring - I believe in it - I'm seriously doubting the insistence
that it's not being steered and no, "survival of the fittest" won't
explain this sort of teleology.

>
> According to creationists, 6,000 years is enough just to produce the
> minor differences we see now between different populations of humans
> globally (600 generations, to stick to the same measure. Fewer if you
> allow for an increasingly delayed sexual maturity in humans). Multiply
> that by almost 1,000. What differences might we then see? Especially
> if you factor in a reproductive barrier, and intense selection
> pressure, not yet experienced by extant groups of humans.
>
> Macro vs micro evolution. How many micros, in your estimation, need to
> be added together to make a macro?
>
> As Chairman Mao was quoted as saying (and I'm probably misquoting,
> though not in essence) 'Even the longest journey starts with a single
> step'. Macro evolution is just loads of micros.- Hide quoted text -
>
However, this "reductionist" claim (Macro = n*Micro) is apparently
hotly debated and unresolved. Nevertheless, before we get lost in
jargon, I would like reiterate the differentiation I am making, in
fact the key question in this post, between a single mutation
producing a slight change (slightly longer tail, ho hum) and a single
mutation producing drastic change (a tail where none was before,
amazing!).

Nic

unread,
May 2, 2007, 6:34:31 AM5/2/07
to
On 1 May, 04:20, "Martin Hutton"
<mdhutton1949REM...@hotmailREMOVE.com> wrote:

Did you mean this one, also by R. Norman?

http://www.talkorigins.org/origins/postmonth/may05.html

Martin Hutton

unread,
May 2, 2007, 11:32:59 PM5/2/07
to

Yep...dat's da bunny. Thx

At least I get the initial if the first name right! :(

--
Martin Hutton

Martin Hutton

unread,
May 2, 2007, 11:30:22 PM5/2/07
to

A single mutation may have no effect (for now) or a massive effect.

> It is my understanding however that only micro-
> evolution (basically adaptation) is well supported by empirical
> evidence and that the time-scales and numbers required for things like

I think Haldane's dilemma has been well answered by more brilliant
minds than mine. You might want to reconsider your thesis.

> chimps to become humans are ludicrously large and do not align with
> the timescales from the fossil record by several orders of magnitude.

First chimps did not become humans...an ancestor population gave
rise to descendant populations of chimps and humans. And, secondly,
who did these calculations and why aren't paleontologists aware of
them?

And then you have to ask...what are the differences between apes
and humans? We have:
1) Bipedal stance: mainly changes in pelvis and location of the
foramen magma (hole in skull for backbone)
2) Face flatness
3) Reduction in canines
4) Reduced body hair
5) Longer brain growth time/childhood

It has been proposed that humans are neotonous* apes...juvenile
apes have many characteristics of humans (more bipedal pelvis,
lower foramen, flat "baby" face, no major canines, reduced body
hair). Thus a simple mutation that changes development schedules
can be a basis for human development including the all important
continued development.

*Neotony: Sexually mature individuals maintaining juvenile
characteristics.

Just something to consider.

--
Martin Hutton

Jack

unread,
May 7, 2007, 4:36:58 PM5/7/07
to
Small correction: Replace "chimp" with "common ancestor". Folklore
dies hard!

Perplexed in Peoria

unread,
May 7, 2007, 5:58:49 PM5/7/07
to

"Jack" <caw...@gmail.com> wrote in message news:1178047226.7...@n59g2000hsh.googlegroups.com...

> On May 1, 7:40 pm, Bob D <ju.d...@btopenworld.com> wrote:
> > On 1 May, 16:06, Jack <cawo...@gmail.com> wrote:
> > > I have the feeling that when you guys talk about a "small change"
> > > we're still talking about hundreds if not thousands of what I call
> > > "stages". I think it would help to clarify what I mean by "stage": I'm
> > > talking about a single genetic mutation. If, as I understand from your
> > > post, a single mutation is _generally_ able to produce dramatic
> > > effects then we have a good case for macro-evolution: evolution
> > > proceeding in jumps, seemingly unlikely but counter-balanced by sheer
> > > numbers and time scale. It is my understanding however that only micro-
> > > evolution (basically adaptation) is well supported by empirical
> > > evidence and that the time-scales and numbers required for things like
> > > chimps to become humans are ludicrously large and do not align with
> > > the timescales from the fossil record by several orders of magnitude.
> >
> > So 5 million years is, what, too long? Too short? 500,000 generations
> > is too few for what kind of changes (assuming 10 years to sexual
> > maturity) required from a common ancestor with, say, a chimp?
>
> Apparently it is (too short that is) and the math has been done. In
> fact if you assume a 3% difference between the chimp and human genomes
> (both not fully resolved to date) corresponding to 90 million base
> pairs you start to see the challenge. Looking only at the active part
> of the genome this reduces to by a further factor of 1%, or aprox. 1
> million base pairs. Looks like we need a directional mutation every 2
> generations.

I don't know what you mean by 'directional', but a mutation roughly every
two generations in what you call the 'active part of the genome' is about
right.

> However, looking at the maximum measured rates of
> mutation in cell division (1 in 10) or in the more robust gametes (1
> in 10'000 - 100'000) the challenge becomes almost insurmountable.

I have no idea where you got the absurd numbers of the preceding sentence.
For humans, something like 50 point mutations per haploid genome per
generation is the usual estimate.

> Now
> you have to start assuming these mutations accumulate in a certain
> direction (i.e. are survival-relevant), do not get lost or reversed by
> later mutations and you start to see the miraculousness of us evolving
> from chimps.

Those things are all pretty much irrelevant. For example, if a random
2% of the base pairs change between man and chimp, then the expectation
is that only 0.04% of them will have changed twice. It doesn't make
much difference. Furthermore, we expect about 50 mutations per generation
in *every* lineage. Selecting out only those lineages lucky enough to
get the 'good' mutations doesn't make much difference at all in the count
of mutations.

> Disclaimer: I'm not trying to cast doubt on evolution
> occuring - I believe in it - I'm seriously doubting the insistence
> that it's not being steered and no, "survival of the fittest" won't
> explain this sort of teleology.

Earlier, you wrote "the math has been done". Well, I have to say that
wherever you got the math, it was from someone who WAS trying to cast
doubt on evolution. Please be more careful in believing the math of
people with a history of lying.

John Harshman

unread,
May 7, 2007, 6:17:09 PM5/7/07
to
Perplexed in Peoria wrote:

Also, the difference between the chimp and human genomes is only 1.2%
(35 million mutations), of which about half is in each lineage. So the
actual percentage is about 0.6%, or a bit under 18 million. If you add
in non-point mutations, the total in both lineages is around 40 million,
20 million each. Since I have no idea what "the active part" means, I
can't determine what that number would be.

Certainly the math he shows here makes no sense, being based on wrong
numbers and false premises. You didn't make this clear enough, so I'll
repeat it: the overwhelming majority of differences between humans and
chimps are neutral, and have nothing to do with selection. The rate of
neutral evolution in a population, conveniently, is the same as the
mutation rate for an individual.

Nic

unread,
May 7, 2007, 7:36:37 PM5/7/07
to
On 7 May, 23:17, John Harshman <jharshman.diespam...@pacbell.net>
wrote:
> Perplexed in Peoria wrote:
> > "Jack" <cawo...@gmail.com> wrote in messagenews:1178047226.7...@n59g2000hsh.googlegroups.com...

As you say further on, the overwhelming majority are neutral, which I
guess is why you split the 40 million in half.

The selected-for ones however don't appear to be equally shared
between the 2 lineages. I'm not sure how safe these recently reported
conclusions (now that rhesus is being used as an outgroup) will prove
to be

http://www.the-scientist.com/news/display/53076/

but if they stand up, it looks as if *selected-for*, *coding*
sequences have changed more in the chimps.

I assume humans may make up for this in selected-for regulatory
sequence changes, particularly ones for brain development.

Jack

unread,
May 9, 2007, 6:26:30 AM5/9/07
to
> > Disclaimer: I'm not trying to cast doubt on evolution
> > occuring - I believe in it - I'm seriously doubting the insistence
> > that it's not being steered and no, "survival of the fittest" won't
> > explain this sort of teleology.
>
> Earlier, you wrote "the math has been done". Well, I have to say that
> wherever you got the math, it was from someone who WAS trying to cast
> doubt on evolution. Please be more careful in believing the math of
> people with a history of lying.

No, and please don't jump to conclusions that the world is divided
into Evolutionists and Creationists. Evolution (the phenomena, not the
pesudo-religion) as such, does not make the the addendum claim "and
all this is blind inevitability". This controversal addendum is
neither scientifically sound nor necessary. In short, Evolution is
just another scientific model which increases our understanding of how
creation works. Gravity or Thermodynamics did not disprove God so why
should evolution?

Jack

unread,
May 9, 2007, 6:29:37 AM5/9/07
to
On May 7, 11:58 pm, "Perplexed in Peoria" <jimmene...@sbcglobal.net>
wrote:
> "Jack" <cawo...@gmail.com> wrote in messagenews:1178047226.7...@n59g2000hsh.googlegroups.com...

I invite you to cite this claim. Here are my sources:

1) The high fidelity of DNA duplication.
M Radman, R Wagner - Sci Am, 1988 - ncbi.nlm.nih.gov

2) Population Genetics and Evolution
LE Mettler, TG Gregg - 1969 - Prentice-Hall

Perplexed in Peoria

unread,
May 9, 2007, 11:18:35 AM5/9/07
to

"Jack" <caw...@gmail.com> wrote in message news:1178706390.0...@n59g2000hsh.googlegroups.com...

Both words, 'evolution' and 'creation' mean different things to different
people. From what you write above, it appears that you are what is often
called a 'theistic evolutionist'. You accept the idea of a long history
of change in the biosphere and its inhabitants, but you suspect that the
Deity had a hand in intimately guiding the process. Feel free to correct
me if I am guessing wrong.

You attack the 'pseudo-religion' of evolution, but I don't think you
understand it. For example, almost no one believes that "all this is
blind inevitability". That phrase doesn't even make sense. But I and
my fellow 'pseudo-religionists' might recognize that the alternative
caricature "all this is the result of blind chance" is directed at us,
and react with annoyance.

In your earlier posting, you sought to show that the number of differences
between man and chimp could not have been the result of random mutation
and provides evidence of guidance. I am quite sure you are wrong about
this. The math works quite well for the overwhelming bulk of those changes
to have been the result of chance mutations. We can't exclude the possibility
that some important and crucial fraction of those mutations were guided,
and we probably will never be able to exclude that possibility. But you
are just wrong that there is evidence for theisticly guided evolution in
the numbers available to us now. Even TE-oriented experts like Francis Collins
wouldn't disagree with me on this.

Perplexed in Peoria

unread,
May 9, 2007, 12:03:45 PM5/9/07
to

"Jack" <caw...@gmail.com> wrote in message news:1178706577.0...@p77g2000hsh.googlegroups.com...
http://www.pnas.org/cgi/reprint/022629899v1.pdf

Says in the abstract:
Our results suggest that the average mammalian genome
mutation rate is 2.2 x 10^-9 per base pair per year

Which, assuming a genome size of 3.5 x 10^9 base pairs
and a generation time of twenty years works out to
154 mutations per genome per generation. About three
times as large as my estimate. Right ballpark, anyways.
But that is mammals in general rather than humans. And
it could be rightly pointed out that this is circular
reasoning since them are getting these rates from genome
comparisons and ASSUMING neutral evolution from common
ancestors.

Here is a paper which does a bit better in this regard:

http://www.genetics.org/cgi/content/full/156/1/297

Says in the abstract:
The average mutation rate was estimated to be 2.5 x 10^-8
mutations per nucleotide site or 175 mutations per diploid
genome per generation.

> Here are my sources:
>
> 1) The high fidelity of DNA duplication.
> M Radman, R Wagner - Sci Am, 1988 - ncbi.nlm.nih.gov
>
> 2) Population Genetics and Evolution
> LE Mettler, TG Gregg - 1969 - Prentice-Hall

But your sources are not available online, so I can't check them.
And you don't quote them, so I have no idea how and whether you
are misinterpreting. '1 in 10' what? In what species? You say
that the challenge of reconciling the numbers is 'almost
insurmountable'. I have to admit that I am finding it quite
difficult to reconcile them since I don't even know what your
numbers are supposed to mean!

Jack

unread,
May 10, 2007, 7:02:19 AM5/10/07
to
On May 9, 5:18 pm, "Perplexed in Peoria" <jimmene...@sbcglobal.net>
wrote:
> "Jack" <cawo...@gmail.com> wrote in messagenews:1178706390.0...@n59g2000hsh.googlegroups.com...

> > > > Disclaimer: I'm not trying to cast doubt on evolution
> > > > occuring - I believe in it - I'm seriously doubting the insistence
> > > > that it's not being steered and no, "survival of the fittest" won't
> > > > explain this sort of teleology.
>
> > > Earlier, you wrote "the math has been done". Well, I have to say that
> > > wherever you got the math, it was from someone who WAS trying to cast
> > > doubt on evolution. Please be more careful in believing the math of
> > > people with a history of lying.
>
> > No, and please don't jump to conclusions that the world is divided
> > into Evolutionists and Creationists. Evolution (the phenomena, not the
> > pesudo-religion) as such, does not make the the addendum claim "and
> > all this is blind inevitability". This controversal addendum is
> > neither scientifically sound nor necessary. In short, Evolution is
> > just another scientific model which increases our understanding of how
> > creation works. Gravity or Thermodynamics did not disprove God so why
> > should evolution?
>
> Both words, 'evolution' and 'creation' mean different things to different
> people. From what you write above, it appears that you are what is often
> called a 'theistic evolutionist'. You accept the idea of a long history
> of change in the biosphere and its inhabitants, but you suspect that the
> Deity had a hand in intimately guiding the process. Feel free to correct
> me if I am guessing wrong.
You are mostly correct. One point though on which I have not
established
an opinion is whether or not God is directly and intimately involved
in the natural
workings of creation. Of course, God is behind it (nature) all, but
whether he guides
it or just fine tuned (incredibly) the start parameters and set off
the Big Bang looks
to me to be speculation and meta-physics.

By the way: It's not the addendum itself that I am attacking. I am
just "reminding the user" that this is no longer science but pseudo-
science. Perhaps you could explain what you call your pseudo-religous
addendum to the Evolution theory? I would be very interested to see if
there is any variation on the "this all happens without the need for
God" addendum.

0 new messages