Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Re: Turtle genome sequence and analysis
79 views
Skip to first unread message
pnyikos
Mar 29, 2013, 12:26:08 PM3/29/13
to nyi...@bellsouth.net
On Mar 29, 7:04锟絘m, Ron O <rokim...@cox.net> wrote:
> Another interesting genome has been published. 锟絀t comes from the same
> lineage as crocs, dinos and birds.
This has been disputed by Gauthier and others, based on fossil
evidence. They even dispute the claim that turtles are secondarily,
rather than primarily, anapsid. If they are right, turtles are the
outliers of extant Sauropsida.
> http://genomebiology.com/2013/14/3/R28/abstract
I looked at the available PDF of the whole provisional article, and a
word search turns up no instances of the word "Sphenodon." [I even
searched for the letter string "Rhync".]
This is a serious omission. Fossil evidence has generally been
interpreted as having Sphenodon as the sister group of archosaurs
among living taxa.
This seems to be a recurring theme -- fossil evidence leading to
opposite conclusions than genomic evidence. We saw it in the case
where genomic evidence leads researchers to place the LCA of living
placental (as opposed to marsupial, etc) mammals at ca. 105 mya, while
fossil evidence was interpreted by them to place it past the K-T
extinction event.
For more details, see:
http://groups.google.com/group/sci.bio.paleontology/msg/d0ec8fc6d5deed91
> The article is open access, but I wouldn't print it out. 锟絀n
> prepublication form it is over 70 pages.
>
> This turtle's genome appears to evolve more slowly than mammal
> genomes. 锟絋hey claim a rate 1/3 of that of humans.
I wonder how they established this, and what the evidence would show
as the rate of evolution of Sphenodon. Osteologically it is a "living
fossil," hardly changed since Jurassic times.
>锟紸lso of interest
> to this group is that they looked at the genes associated with teeth
> (turtles lost their teeth) and found the same genes degenerating in
> the genome as are found in birds. 锟組ore evidence of common descent and
> inheriting a genome from an ancestor with teeth.
If those same genes are found in Sphenodon, that could be significant.
> 锟絃osing teeth is a
> tricky proposition. 锟結ou have to maintain the cranial structure of the
> ancestral type, but get rid of the teeth.
I can see that where *gaining* teeth is concerned, but why *losing*
teeth?
> So you need to keep some
> aspects of developing teeth, while changing others.
I still don't get it.
> These kinds of finds will keep coming. 锟絋he 10,000 animal genome
> project has turned into 100,000 so over the next few years there is
> going to be so much genomic sequence available that we could probably
> take a century analyzing it.
>
> Some researchers are already doing population scale genome analysis,
> not just individual genomes of individual species. 锟絋he 1,000 human
> genome project turned into the 10,000 human genome project and they
> have pretty much covered the first thousand and are working on the
> 10,000.
I've read in one place, though, that the whole human genome has not
actually been sequenced, because e.g. some junk DNA remains
unsequenced. Do you know anything about this?
Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer--
University of South Carolina
http://www.math.sc.edu/~nyikos/
nyikos @ math.sc.edu
> Another interesting genome has been published. 锟絀t comes from the same
> lineage as crocs, dinos and birds.
This has been disputed by Gauthier and others, based on fossil
evidence. They even dispute the claim that turtles are secondarily,
rather than primarily, anapsid. If they are right, turtles are the
outliers of extant Sauropsida.
> http://genomebiology.com/2013/14/3/R28/abstract
I looked at the available PDF of the whole provisional article, and a
word search turns up no instances of the word "Sphenodon." [I even
searched for the letter string "Rhync".]
This is a serious omission. Fossil evidence has generally been
interpreted as having Sphenodon as the sister group of archosaurs
among living taxa.
This seems to be a recurring theme -- fossil evidence leading to
opposite conclusions than genomic evidence. We saw it in the case
where genomic evidence leads researchers to place the LCA of living
placental (as opposed to marsupial, etc) mammals at ca. 105 mya, while
fossil evidence was interpreted by them to place it past the K-T
extinction event.
For more details, see:
http://groups.google.com/group/sci.bio.paleontology/msg/d0ec8fc6d5deed91
> The article is open access, but I wouldn't print it out. 锟絀n
> prepublication form it is over 70 pages.
>
> This turtle's genome appears to evolve more slowly than mammal
> genomes. 锟絋hey claim a rate 1/3 of that of humans.
I wonder how they established this, and what the evidence would show
as the rate of evolution of Sphenodon. Osteologically it is a "living
fossil," hardly changed since Jurassic times.
>锟紸lso of interest
> to this group is that they looked at the genes associated with teeth
> (turtles lost their teeth) and found the same genes degenerating in
> the genome as are found in birds. 锟組ore evidence of common descent and
> inheriting a genome from an ancestor with teeth.
If those same genes are found in Sphenodon, that could be significant.
> 锟絃osing teeth is a
> tricky proposition. 锟結ou have to maintain the cranial structure of the
> ancestral type, but get rid of the teeth.
I can see that where *gaining* teeth is concerned, but why *losing*
teeth?
> So you need to keep some
> aspects of developing teeth, while changing others.
I still don't get it.
> These kinds of finds will keep coming. 锟絋he 10,000 animal genome
> project has turned into 100,000 so over the next few years there is
> going to be so much genomic sequence available that we could probably
> take a century analyzing it.
>
> Some researchers are already doing population scale genome analysis,
> not just individual genomes of individual species. 锟絋he 1,000 human
> genome project turned into the 10,000 human genome project and they
> have pretty much covered the first thousand and are working on the
> 10,000.
I've read in one place, though, that the whole human genome has not
actually been sequenced, because e.g. some junk DNA remains
unsequenced. Do you know anything about this?
Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer--
University of South Carolina
http://www.math.sc.edu/~nyikos/
nyikos @ math.sc.edu
Ron O
Mar 29, 2013, 5:37:00 PM3/29/13
to
On Mar 29, 11:26�am, pnyikos <nyik...@bellsouth.net> wrote:
> On Mar 29, 7:04�am, Ron O <rokim...@cox.net> wrote:
Face it Nyikos about the only post that I want to see from you is your
third self inflicted knockdown post so that you will just leave me
alone. The knockdown never appeared in February like you claimed that
it would or in January like you claimed it would or soon after your
second stupidity last October.
What makes you think that I am at all interested in what you have to
post after your last assoholic episode? What a loser.
Ron Okimoto
> On Mar 29, 7:04�am, Ron O <rokim...@cox.net> wrote:
Face it Nyikos about the only post that I want to see from you is your
third self inflicted knockdown post so that you will just leave me
alone. The knockdown never appeared in February like you claimed that
it would or in January like you claimed it would or soon after your
second stupidity last October.
What makes you think that I am at all interested in what you have to
post after your last assoholic episode? What a loser.
Ron Okimoto
Thrinaxodon
Mar 29, 2013, 6:52:11 PM3/29/13
to
On Friday, March 29, 2013 5:37:00 PM UTC-4, Ron O wrote:
> On Mar 29, 11:26 am, pnyikos <nyik...@bellsouth.net> wrote:
>
> > On Mar 29, 7:04 am, Ron O <rokim...@cox.net> wrote:
>
>
>
> Face it Nyikos about the only post that I want to see from you is your
>
> third self inflicted knockdown post so that you will just leave me
>
> alone. The knockdown never appeared in February like you claimed that
>
> it would or in January like you claimed it would or soon after your
>
> second stupidity last October.
>
> > It most likely won't happen. He's already bitch slapped me by email. He'll most likely, won't leave you alone either.
> On Mar 29, 11:26 am, pnyikos <nyik...@bellsouth.net> wrote:
>
> > On Mar 29, 7:04 am, Ron O <rokim...@cox.net> wrote:
>
>
>
> Face it Nyikos about the only post that I want to see from you is your
>
> third self inflicted knockdown post so that you will just leave me
>
> alone. The knockdown never appeared in February like you claimed that
>
> it would or in January like you claimed it would or soon after your
>
> second stupidity last October.
>
John Harshman
Mar 29, 2013, 7:37:53 PM3/29/13
to
should have >. That is, your comment is wrongly attributed to Ron Okimoto.
pnyikos
Apr 4, 2013, 2:54:36 PM4/4/13
to nyi...@math.sc.edu
On Mar 29, 12:47 pm, John Harshman <jharsh...@pacbell.net> wrote:
> On 3/29/13 9:26 AM, pnyikos wrote:
>
> > On Mar 29, 7:04 am, Ron O<rokim...@cox.net> wrote:
> >> Another interesting genome has been published. It comes from the same
> past few years. Turtles are either archosaurs of the living sister group
> of archosaurs.
>
> There is of course some morphological evidence that they are at least
> diapsids, mostly advanced by Olivier Rieppel and colleagues.
Not here, unless one of the co-authors here is a colleague:
> >>http://genomebiology.com/2013/14/3/R28/abstract
Are you referring to the papers listed in the following passage?
"The position of turtles (Testudines) is uncertain; some authors place
them approximately in the position shown above (Laurin & Reisz, 1995;
Lee, 1993, 1995, 2001; Frost et al., 2006; Werneburg & Sánchez-
Villagra, 2009; Lyson et al. 2010), while others place them among
Diapsida (deBraga & Rieppel, 1996, 1997; Rieppel & Reisz, 1999; Hedges
& Poling, 1999; Mannen & Li, 1999; Hugall et al., 2007; Li et al.,
2008)."
-- caption to phylogenetic tree of Amniota in
http://tolweb.org/Amniota/14990
The tree shows turtles (Testudines) in Anapsida.
Laurin and Gauthier, who did the site, see your Rieppel et.al. ...1999
and raise it to 2010.
Later down in the page they write:
"At least two total evidence analyses (in which molecular and
morphological data are combined) suggest that turtles are not diapsids
(Lee, 2001; Frost et al., 2006). In both, the molecular characters are
much more numerous than the morphological ones, which implies
significant support for the placement of turtles outside diapsids in
these molecular datasets."
I wonder who is cherry-picking this time, you or they.
> > I looked at the available PDF of the whole provisional article, and a
> > word search turns up no instances of the word "Sphenodon." [I even
> > searched for the letter string "Rhync".]
>
> > This is a serious omission. Fossil evidence has generally been
> > interpreted as having Sphenodon as the sister group of archosaurs
> > among living taxa.
>
> Are you sure? Usually it's interpreted as the sister of squamates.
You are correct, I remembered those old "bubble diagrams"
incorrectly.
However, if Sphenodon evolved much more slowly (molecularlty speaking)
than either birds, crocs, or squamates, then a traditional systematist
would say it is much more closely related to the LCA of the whole
diapsid clade than any of them.
Then a race would be on to see whether tortoises or Sphenodon are
more closely related to the LCA.
This is one race where the slower is the winner. No hares need
apply.
Continued in next post.
> On 3/29/13 9:26 AM, pnyikos wrote:
>
> > On Mar 29, 7:04 am, Ron O<rokim...@cox.net> wrote:
> >> Another interesting genome has been published. It comes from the same
> >> lineage as crocs, dinos and birds.
>
> > This has been disputed by Gauthier and others, based on fossil
> > evidence. They even dispute the claim that turtles are secondarily,
> > rather than primarily, anapsid. If they are right, turtles are the
> > outliers of extant Sauropsida.
>
> If. The molecular data have been getting more and more conclusive in the
>
> > This has been disputed by Gauthier and others, based on fossil
> > evidence. They even dispute the claim that turtles are secondarily,
> > rather than primarily, anapsid. If they are right, turtles are the
> > outliers of extant Sauropsida.
>
> past few years. Turtles are either archosaurs of the living sister group
> of archosaurs.
>
> There is of course some morphological evidence that they are at least
> diapsids, mostly advanced by Olivier Rieppel and colleagues.
Not here, unless one of the co-authors here is a colleague:
> >>http://genomebiology.com/2013/14/3/R28/abstract
Are you referring to the papers listed in the following passage?
"The position of turtles (Testudines) is uncertain; some authors place
them approximately in the position shown above (Laurin & Reisz, 1995;
Lee, 1993, 1995, 2001; Frost et al., 2006; Werneburg & Sánchez-
Villagra, 2009; Lyson et al. 2010), while others place them among
Diapsida (deBraga & Rieppel, 1996, 1997; Rieppel & Reisz, 1999; Hedges
& Poling, 1999; Mannen & Li, 1999; Hugall et al., 2007; Li et al.,
2008)."
-- caption to phylogenetic tree of Amniota in
http://tolweb.org/Amniota/14990
The tree shows turtles (Testudines) in Anapsida.
Laurin and Gauthier, who did the site, see your Rieppel et.al. ...1999
and raise it to 2010.
Later down in the page they write:
"At least two total evidence analyses (in which molecular and
morphological data are combined) suggest that turtles are not diapsids
(Lee, 2001; Frost et al., 2006). In both, the molecular characters are
much more numerous than the morphological ones, which implies
significant support for the placement of turtles outside diapsids in
these molecular datasets."
I wonder who is cherry-picking this time, you or they.
> > I looked at the available PDF of the whole provisional article, and a
> > word search turns up no instances of the word "Sphenodon." [I even
> > searched for the letter string "Rhync".]
>
> > This is a serious omission. Fossil evidence has generally been
> > interpreted as having Sphenodon as the sister group of archosaurs
> > among living taxa.
>
You are correct, I remembered those old "bubble diagrams"
incorrectly.
However, if Sphenodon evolved much more slowly (molecularlty speaking)
than either birds, crocs, or squamates, then a traditional systematist
would say it is much more closely related to the LCA of the whole
diapsid clade than any of them.
Then a race would be on to see whether tortoises or Sphenodon are
more closely related to the LCA.
This is one race where the slower is the winner. No hares need
apply.
Continued in next post.
pnyikos
Apr 4, 2013, 3:20:23 PM4/4/13
to nyi...@math.sc.edu
On Mar 29, 12:47�pm, John Harshman <jharsh...@pacbell.net> wrote:
> On 3/29/13 9:26 AM, pnyikos wrote:
>
> > On Mar 29, 7:04 am, Ron O<rokim...@cox.net> �wrote:
> On 3/29/13 9:26 AM, pnyikos wrote:
>
> > On Mar 29, 7:04 am, Ron O<rokim...@cox.net> �wrote:
> > This seems to be a recurring theme -- fossil evidence leading to
> > opposite conclusions than genomic evidence. �We saw it in the case
> > where genomic evidence leads researchers to place the LCA of living
> > placental (as opposed to marsupial, etc) mammals at ca. 105 mya, while
> > fossil evidence was interpreted by them to place it past the K-T
> > extinction event.
>
> It's a recurring theme only because you (and the world generally) take
> > opposite conclusions than genomic evidence. �We saw it in the case
> > where genomic evidence leads researchers to place the LCA of living
> > placental (as opposed to marsupial, etc) mammals at ca. 105 mya, while
> > fossil evidence was interpreted by them to place it past the K-T
> > extinction event.
>
> note of the times it happens but not of the much more frequent times it
> doesn't.
Either that, or there are many other times having to do with closely
related taxa, and so they escape notice. [For cladophiles like you, I
should write "molecularly and morphologically similar" rather than
"closely related" so as to make it intelligible to you.]
> Man bites dog.
Or maybe "cats bite hundreds of dogs, but it is only when man gets
into the act that news reporters take notice."
> > For more details, see:
>
> >http://groups.google.com/group/sci.bio.paleontology/msg/d0ec8fc6d5deed91
>
> >> prepublication form it is over 70 pages.
>
> >> This turtle's genome appears to evolve more slowly than mammal
> >> genomes. �They claim a rate 1/3 of that of humans.
>
> >> This turtle's genome appears to evolve more slowly than mammal
>
> > I wonder how they established this, and what the evidence would show
> > as the rate of evolution of Sphenodon. Osteologically it is a "living
> > fossil," hardly changed since Jurassic times.
>
> There is no apparent correlation, based on all sorts of data in all
> > I wonder how they established this, and what the evidence would show
> > as the rate of evolution of Sphenodon. Osteologically it is a "living
> > fossil," hardly changed since Jurassic times.
>
> sorts of taxa, between morphological and molecular evolutionary rates.
However, I am still curious to know what the situation is with
Sphenodon, as suggested in my first reply.
> "How they established this" is presumably either by comparison of both
> with outgroups ("relative rate test") or by using time-calibrated
> phylogenies.
The details would be nice to know.
> >> � Also of interest
> >> to this group is that they looked at the genes associated with teeth
> >> (turtles lost their teeth) and found the same genes degenerating in
> >> the genome as are found in birds. �More evidence of common descent and
> >> (turtles lost their teeth) and found the same genes degenerating in
> >> inheriting a genome from an ancestor with teeth.
Yet another of millions of little nails in the coffin of creationism.
> > If those same genes are found in Sphenodon, that could be significant.
>
> both bird and turtle ancestors had teeth.
Well, duh. I was interested in how much light the same genes in
Sphenodon shed on what the LCA of extant Sauropsida was like.
> >> � Losing teeth is a
> >> tricky proposition. �You have to maintain the cranial structure of the
> >> ancestral type, but get rid of the teeth.
>
> > I can see that where *gaining* teeth is concerned, but why *losing*
> > teeth?
>
> Not clear to me either. The presumption here would be that developing
>
> > I can see that where *gaining* teeth is concerned, but why *losing*
> > teeth?
>
> teeth are crucial in the development of other parts of the skull. That
> seems an odd proposition.
Ron O doesn't want to discuss this with me, but I was hoping he would
discuss it with you. But so far, it's no go.
Maybe if you replied to him directly...
Peter Nyikos
pnyikos
Apr 4, 2013, 3:26:24 PM4/4/13
to nyi...@bellsouth.net
On Mar 29, 5:37 pm, Ron O <rokim...@cox.net> neglected to snip the
following attributions, but snipped all the text from the posts to
which they refer.
> On Mar 29, 11:26 am, pnyikos <nyik...@bellsouth.net> wrote:
>
> > On Mar 29, 7:04 am, Ron O <rokim...@cox.net> wrote:
By some people's standards [but not mine], Ron O replied to himself:
> Face it Nyikos about the only post that I want to see from you
would be off topic, and so nothing on the same topic is going to be
posted by me here.
Post your off-topic taunts on some thread where they are no more off-
topic than some of the surrounding posts.
[snip rest of off-topic taunt]
Note to others: if anyone besides Ron O thinks what I snipped is
worthy of being addressed, I'll address it on a more appropriate
thread. Otherwise I'm finished with it.
For the benefit of these hypothetical readers, the Ron O post to
which I am replying can be found here:
http://groups.google.com/group/talk.origins/msg/98b5a70fff8f9a50
Peter Nyikos
following attributions, but snipped all the text from the posts to
which they refer.
> On Mar 29, 11:26 am, pnyikos <nyik...@bellsouth.net> wrote:
>
> > On Mar 29, 7:04 am, Ron O <rokim...@cox.net> wrote:
> Face it Nyikos about the only post that I want to see from you
posted by me here.
Post your off-topic taunts on some thread where they are no more off-
topic than some of the surrounding posts.
[snip rest of off-topic taunt]
Note to others: if anyone besides Ron O thinks what I snipped is
worthy of being addressed, I'll address it on a more appropriate
thread. Otherwise I'm finished with it.
For the benefit of these hypothetical readers, the Ron O post to
which I am replying can be found here:
http://groups.google.com/group/talk.origins/msg/98b5a70fff8f9a50
Peter Nyikos
Ron O
Apr 4, 2013, 7:04:27 PM4/4/13
to
On Apr 4, 2:26�pm, pnyikos <nyik...@bellsouth.net> wrote:
> On Mar 29, 5:37�pm, Ron O <rokim...@cox.net> neglected to snip the
> following attributions, but snipped all the text from the posts to
> which they refer.
>
> > On Mar 29, 11:26 am, pnyikos <nyik...@bellsouth.net> wrote:
>
> > > On Mar 29, 7:04 am, Ron O <rokim...@cox.net> wrote:
>
> By some people's �standards [but not mine], Ron O replied to himself:
>
> > Face it Nyikos about the only post that I want to see from you
>
> would be off topic, and so nothing on the same topic is going to be
> posted by me here.
How delusional can a person get? Nyikos is the king of posting stupid
> On Mar 29, 5:37�pm, Ron O <rokim...@cox.net> neglected to snip the
> following attributions, but snipped all the text from the posts to
> which they refer.
>
> > On Mar 29, 11:26 am, pnyikos <nyik...@bellsouth.net> wrote:
>
> > > On Mar 29, 7:04 am, Ron O <rokim...@cox.net> wrote:
>
> By some people's �standards [but not mine], Ron O replied to himself:
>
> > Face it Nyikos about the only post that I want to see from you
>
> would be off topic, and so nothing on the same topic is going to be
> posted by me here.
junk to other posters in other threads, and he balks at what? Doing
what he claimed that he would do nearly a year and a half ago.
My post was so short that the only reason why Nyikos would snip any of
it out is to cowardly run away from reality again.
QUOTE:
Face it Nyikos about the only post that I want to see from you is your
third self inflicted knockdown post so that you will just leave me
alone. The knockdown never appeared in February like you claimed that
it would or in January like you claimed it would or soon after your
second stupidity last October.
What makes you think that I am at all interested in what you have to
post after your last assoholic episode? What a loser.
END QUOTE:
third self inflicted knockdown post so that you will just leave me
alone. The knockdown never appeared in February like you claimed that
it would or in January like you claimed it would or soon after your
second stupidity last October.
What makes you think that I am at all interested in what you have to
post after your last assoholic episode? What a loser.
Really, after all the bull shit that Nyikos is guilty of over more
than 2 years he has to claim that he is unwilling to post something
that is off topic? Who changed the name of the thread and started a
totally off topic side thread and claimed it as his second knockdown?
That was Nyikos. The guy that just made up the stupid story that he
claimed was so funny in that side thread. You can't make this junk
up. Nyikos will always do himself one better in the bogousity
department.
>
> Post your off-topic taunts on some thread where they are no more off-
> topic than some of the surrounding posts.
your opinion on anything? No.
Now all you can do is run. How sad is that? Who has been claiming
that the hammer blows and knockdowns are coming for nearly a year and
a half? Who has been the pathetic liar and pretender in all that
time?
>
> [snip rest of off-topic taunt]
>
> Note to others: if anyone besides Ron O thinks what I snipped is
> worthy of being addressed, I'll address it on a more appropriate
> thread. Otherwise I'm finished with it.
cowardly lying asshole. The sad thing is that Nyikos knows what he is
and has to run from reality. Who has been bleating and lying about me
to other posters? Who can't face reality and has to run and bleat to
everyone else?
>
> For the benefit of these hypothetical readers, �the Ron O post to
> which I am replying can be found here:
>
> http://groups.google.com/group/talk.origins/msg/98b5a70fff8f9a50
>
> Peter Nyikos
good. Why keep coming back to whine and bleat about bogus junk that
you did years ago? Just do what you bragged about doing over a year
ago and then don't darken my door again. I will, of course, have to
demonstrate what a whacked out loser you are when you do deliver the
non existent knockdown and punch yourself in the face again. So just
do it and then whine and bleat to your hearts content, just don't
bother posting to me unless you like being reminded what a low life
cowardly lying scum bag you are.
Really, instead of continuing to whine and bleat, just deliver your
third stupid knockdown and run away for good. How many times are you
going to say that you are going to deliver it and then end up lying?
Did it come soon in October? Did it come in January like you said
that it would? Did it come in February like you said that it would?
What kind of whining and bleating were you doing instead of delivering
your stupid knockdown for the last couple of months? Why shouldn't I
be ticked off? You are such a degenerate loser and all you do is keep
coming back with your stupid junk time and time again. Well just
deliver the stupid knockdown and quit like you said that you would.
Just think of it as one last lie that you have to tell. That should
help the pathological liar passed his yellow streak.
Nyikos can call this taunting, but he knows for a fact that it is all
true. So just deliver your third stupid knockdown and stop posting to
me like you said that you would over a year ago.
Ron Okimoto
pnyikos
Apr 10, 2013, 12:07:03 PM4/10/13
to nyi...@bellsouth.net
John, this thread has been pretty well hijacked in talk.origins by
people making personal attacks, so I would like for us to move this on-
topic discussion to sci.bio.paleontology. There, the majority of
posts to this thread are still completely on-topic.
Accordingly, I've added s.b.p. to the newsgroups and will be making
on-topic comments after the attributions and earlier comments.
I'll see if I can interest Richard Norman in joining us over there.
His knowledge of paleontology could be of help.
On Apr 4, 10:50�pm, John Harshman <jharsh...@pacbell.net> wrote:
> >> of archosaurs.
Still, Gauthier et. al. were holding out last year, according to the
article whose abstract Ron O linked:
21. Lyson TR, Sperling EA, Heimberg AM, Gauthier JA, King BL, Peterson
KJ:
MicroRNAs support a turtle plus lizard clade. Biol Lett 2012,
8:104-107.
-- http://genomebiology.com/content/pdf/gb-2013-14-3-r28.pdf
See also a comment about two total analyses from their webpage [posted
below]. But back to your last comment...
Is there actually molecular support for turtles being more closely
related to birds (in the cladistic sense) than to crocs??? That's
what it would take for molecular data to show that turtles are
archosaurs.
At the opposite extreme, as long as only fragments of the genome of
*Sphenodon* have been sequenced, it is too early to tell whether
turtles are the living sister group of archosaurs, even from a
molecular viewpoint.
> >> There is of course some morphological evidence that they are at least
> >> diapsids, mostly advanced by Olivier Rieppel and colleagues.
>
> > Not here, unless one of the co-authors here is a colleague:
>
> >>>>http://genomebiology.com/2013/14/3/R28/abstract
>
> > Are you referring to the papers listed in the following passage?
>
> > "The position of turtles (Testudines) is uncertain; some authors place
> > them approximately in the position shown above (Laurin& �Reisz, 1995;
> > Lee, 1993, 1995, 2001; Frost et al., 2006; Werneburg& �S nchez-
> > & �Poling, 1999; Mannen& �Li, 1999; Hugall et al., 2007; Li et al.,
Do you have a morphological analysis more recent than 1999 for turtles
being diapsids? Gauthier et. al. have more recent ones to point to,
including total analyses (see below).
> > The tree shows turtles (Testudines) in Anapsida.
>
> Since the various references conflict, they picked one over the others.
Seems they picked several 21st century ones over 20th century ones,
where morphology is concerned.
> > Laurin and Gauthier, who did the site, see your Rieppel et.al. ...1999
> > and raise it to 2010.
>
> > Later down in the page they write:
>
> > � "At least two total evidence analyses (in which molecular and
> > morphological data are combined) suggest that turtles are not diapsids
> > (Lee, 2001; Frost et al., 2006). In both, the molecular characters are
> > much more numerous than the morphological ones, which implies
> > significant support for the placement of turtles outside diapsids in
> > these molecular datasets."
Note the reasoning about the molecular evidence.
> > I wonder who is cherry-picking this time, you or they.
>
> Neither.
Why did you pick references from the 20th century, then? Was Rieppel
just the first author that popped into your mind?
people making personal attacks, so I would like for us to move this on-
topic discussion to sci.bio.paleontology. There, the majority of
posts to this thread are still completely on-topic.
Accordingly, I've added s.b.p. to the newsgroups and will be making
on-topic comments after the attributions and earlier comments.
I'll see if I can interest Richard Norman in joining us over there.
His knowledge of paleontology could be of help.
On Apr 4, 10:50�pm, John Harshman <jharsh...@pacbell.net> wrote:
> On 4/4/13 11:54 AM, pnyikos wrote:
> > On Mar 29, 12:47 pm, John Harshman<jharsh...@pacbell.net> �wrote:
> >> On 3/29/13 9:26 AM, pnyikos wrote:
>
> >>> On Mar 29, 7:04 am, Ron O<rokim...@cox.net> � �wrote:
> >>>> Another interesting genome has been published. �It comes from the same
> >>>> lineage as crocs, dinos and birds.
>
> >>> This has been disputed by Gauthier and others, based on fossil
> >>> evidence. �They even dispute the claim that turtles are secondarily,
> >>> rather than primarily, anapsid. �If they are right, turtles are the
> >>> outliers of extant Sauropsida.
>
> >> If. The molecular data have been getting more and more conclusive in the
> >> past few years. Turtles are either archosaurs or the living sister group
> > On Mar 29, 12:47 pm, John Harshman<jharsh...@pacbell.net> �wrote:
> >> On 3/29/13 9:26 AM, pnyikos wrote:
>
> >>> On Mar 29, 7:04 am, Ron O<rokim...@cox.net> � �wrote:
> >>>> Another interesting genome has been published. �It comes from the same
> >>>> lineage as crocs, dinos and birds.
>
> >>> This has been disputed by Gauthier and others, based on fossil
> >>> evidence. �They even dispute the claim that turtles are secondarily,
> >>> rather than primarily, anapsid. �If they are right, turtles are the
> >>> outliers of extant Sauropsida.
>
> >> If. The molecular data have been getting more and more conclusive in the
> >> of archosaurs.
Still, Gauthier et. al. were holding out last year, according to the
article whose abstract Ron O linked:
21. Lyson TR, Sperling EA, Heimberg AM, Gauthier JA, King BL, Peterson
KJ:
MicroRNAs support a turtle plus lizard clade. Biol Lett 2012,
8:104-107.
-- http://genomebiology.com/content/pdf/gb-2013-14-3-r28.pdf
See also a comment about two total analyses from their webpage [posted
below]. But back to your last comment...
Is there actually molecular support for turtles being more closely
related to birds (in the cladistic sense) than to crocs??? That's
what it would take for molecular data to show that turtles are
archosaurs.
At the opposite extreme, as long as only fragments of the genome of
*Sphenodon* have been sequenced, it is too early to tell whether
turtles are the living sister group of archosaurs, even from a
molecular viewpoint.
> >> There is of course some morphological evidence that they are at least
> >> diapsids, mostly advanced by Olivier Rieppel and colleagues.
>
> > Not here, unless one of the co-authors here is a colleague:
>
> >>>>http://genomebiology.com/2013/14/3/R28/abstract
>
> > Are you referring to the papers listed in the following passage?
>
> > "The position of turtles (Testudines) is uncertain; some authors place
> > Lee, 1993, 1995, 2001; Frost et al., 2006; Werneburg& �S nchez-
> > Villagra, 2009; Lyson et al. 2010), while others place them among
> > Diapsida (deBraga& �Rieppel, 1996, 1997; Rieppel& �Reisz, 1999; Hedges
> > & �Poling, 1999; Mannen& �Li, 1999; Hugall et al., 2007; Li et al.,
> > 2008)."
> > � �-- caption to phylogenetic tree of Amniota in
> > � �http://tolweb.org/Amniota/14990
>
> The ones by Rieppel are among the ones I was referring to.
> > � �-- caption to phylogenetic tree of Amniota in
> > � �http://tolweb.org/Amniota/14990
>
Do you have a morphological analysis more recent than 1999 for turtles
being diapsids? Gauthier et. al. have more recent ones to point to,
including total analyses (see below).
> > The tree shows turtles (Testudines) in Anapsida.
>
Seems they picked several 21st century ones over 20th century ones,
where morphology is concerned.
> > Laurin and Gauthier, who did the site, see your Rieppel et.al. ...1999
> > and raise it to 2010.
>
> > Later down in the page they write:
>
> > � "At least two total evidence analyses (in which molecular and
> > morphological data are combined) suggest that turtles are not diapsids
> > (Lee, 2001; Frost et al., 2006). In both, the molecular characters are
> > much more numerous than the morphological ones, which implies
> > significant support for the placement of turtles outside diapsids in
> > these molecular datasets."
> > I wonder who is cherry-picking this time, you or they.
>
Why did you pick references from the 20th century, then? Was Rieppel
just the first author that popped into your mind?
pnyikos
Apr 10, 2013, 12:36:59 PM4/10/13
to nyi...@bellsouth.net, r_s_n...@comcast.net
CC: Richard Norman, because of the invitation below.
On Apr 4, 8:36锟絧m, Richard Norman <r_s_nor...@comcast.net> wrote:
> On Thu, 4 Apr 2013 16:52:30 -0700 (PDT), Ron O <rokim...@cox.net>
> wrote:
[huge snip]
> >I'm the one that has had enough. 锟絀 still can't understand why anyone
> >would want to read junk that involves Nyikos.
On this thread, it would presumably be the fact that I've kept on-
topic on this thread, except for one reply to Ron O. That continued
today with a reply I did to John Harshman a short while ago. It had
to do with the evidence of where turtles sit in Sauropsida-- diapsid
v. anapsid, archosaur v. sister group of archosaurs v. sister group of
{archosaurs, tuatara} v. Anapsida.
Richard, I was hoping you could join the two of us on this
discussion. With your knowledge of paleontology, I think you can be a
big help in discussing the morphological evidence, at least.
I've added sci.bio.paleontology to the newsgroups because (see my
reply to Harshman) this thread has been pretty well trashed by all the
personal attacks by others, and there we have a good chance of keeping
it on-topic. In fact, I wouldn't mind if you both started posting
just to s.b.p. I'll be watching for you over there.
> > 锟絀 stopped reading most
> >of the crap Nyikos posts 2 years ago.
>
> If you stopped reading it then how come you respond so vociferously
> and at such enormous length? 锟絎hy do you think we tolerate your crap
> about him any better than his rants? 锟絎hy don't both of you just give
> it a rest? 锟紼ven one of you calling it quits would be an enormous
> improvement.
I've done the nearest thing to that which I comfortably can. On the
"By Their Fruits...." thread I've told Hemidactylus about my new
policy. I gave him an almost complete statement of it, as follows:
"from now on, I will confine myself to very
brief direct replies to posts of Ron O,"
--http://groups.google.com/group/talk.origins/msg/5ec655abe34712f9
I carelessly left off a conditional bit ("in reply to me") that I had
put in a much earlier post to Hemidactylus:
"But after the third knockdown, I will reply to almost
nothing Ron O posts in reply to me..."
-- http://groups.google.com/group/talk.origins/msg/d319bac1c0c85623
Note: "almost nothing," not "almost no posts". That could come
later, if enough people become aware of my new policy. Everyone on
this thread seems to be oblivious to it so far, despite my hewing to
it in the one off-topic post I've done to this thread so far.
Peter Nyikos
On Apr 4, 8:36锟絧m, Richard Norman <r_s_nor...@comcast.net> wrote:
> On Thu, 4 Apr 2013 16:52:30 -0700 (PDT), Ron O <rokim...@cox.net>
> wrote:
[huge snip]
> >I'm the one that has had enough. 锟絀 still can't understand why anyone
> >would want to read junk that involves Nyikos.
On this thread, it would presumably be the fact that I've kept on-
topic on this thread, except for one reply to Ron O. That continued
today with a reply I did to John Harshman a short while ago. It had
to do with the evidence of where turtles sit in Sauropsida-- diapsid
v. anapsid, archosaur v. sister group of archosaurs v. sister group of
{archosaurs, tuatara} v. Anapsida.
Richard, I was hoping you could join the two of us on this
discussion. With your knowledge of paleontology, I think you can be a
big help in discussing the morphological evidence, at least.
I've added sci.bio.paleontology to the newsgroups because (see my
reply to Harshman) this thread has been pretty well trashed by all the
personal attacks by others, and there we have a good chance of keeping
it on-topic. In fact, I wouldn't mind if you both started posting
just to s.b.p. I'll be watching for you over there.
> > 锟絀 stopped reading most
> >of the crap Nyikos posts 2 years ago.
>
> If you stopped reading it then how come you respond so vociferously
> and at such enormous length? 锟絎hy do you think we tolerate your crap
> about him any better than his rants? 锟絎hy don't both of you just give
> it a rest? 锟紼ven one of you calling it quits would be an enormous
> improvement.
I've done the nearest thing to that which I comfortably can. On the
"By Their Fruits...." thread I've told Hemidactylus about my new
policy. I gave him an almost complete statement of it, as follows:
"from now on, I will confine myself to very
brief direct replies to posts of Ron O,"
--http://groups.google.com/group/talk.origins/msg/5ec655abe34712f9
I carelessly left off a conditional bit ("in reply to me") that I had
put in a much earlier post to Hemidactylus:
"But after the third knockdown, I will reply to almost
nothing Ron O posts in reply to me..."
-- http://groups.google.com/group/talk.origins/msg/d319bac1c0c85623
Note: "almost nothing," not "almost no posts". That could come
later, if enough people become aware of my new policy. Everyone on
this thread seems to be oblivious to it so far, despite my hewing to
it in the one off-topic post I've done to this thread so far.
Peter Nyikos
jillery
Apr 10, 2013, 12:52:45 PM4/10/13
to
On Wed, 10 Apr 2013 09:07:03 -0700 (PDT), pnyikos
<nyi...@bellsouth.net> wrote:
>John, this thread has been pretty well hijacked in talk.origins by
>people making personal attacks, so I would like for us to move this on-
>topic discussion to sci.bio.paleontology. There, the majority of
>posts to this thread are still completely on-topic.
>
>Accordingly, I've added s.b.p. to the newsgroups and will be making
>on-topic comments after the attributions and earlier comments.
>
>I'll see if I can interest Richard Norman in joining us over there.
>His knowledge of paleontology could be of help.
The above is quite ironic, since it was Ron O. who started this topic
<nyi...@bellsouth.net> wrote:
>John, this thread has been pretty well hijacked in talk.origins by
>people making personal attacks, so I would like for us to move this on-
>topic discussion to sci.bio.paleontology. There, the majority of
>posts to this thread are still completely on-topic.
>
>Accordingly, I've added s.b.p. to the newsgroups and will be making
>on-topic comments after the attributions and earlier comments.
>
>I'll see if I can interest Richard Norman in joining us over there.
>His knowledge of paleontology could be of help.
in T.O. Which make you the hijacker.
[...]
Richard Norman
Apr 10, 2013, 1:39:14 PM4/10/13
to
>John, this thread has been pretty well hijacked in talk.origins by
>people making personal attacks, so I would like for us to move this on-
>topic discussion to sci.bio.paleontology. There, the majority of
>posts to this thread are still completely on-topic.
>
>Accordingly, I've added s.b.p. to the newsgroups and will be making
>on-topic comments after the attributions and earlier comments.
>
>I'll see if I can interest Richard Norman in joining us over there.
>His knowledge of paleontology could be of help.
>
<snip on topic discussion>
>people making personal attacks, so I would like for us to move this on-
>topic discussion to sci.bio.paleontology. There, the majority of
>posts to this thread are still completely on-topic.
>
>Accordingly, I've added s.b.p. to the newsgroups and will be making
>on-topic comments after the attributions and earlier comments.
>
>I'll see if I can interest Richard Norman in joining us over there.
>His knowledge of paleontology could be of help.
>
I actually have virtually no knowledge of paleontology. I am just an
interested observer with some knowledge of biology and the ability to
read the literature with a bit of understanding.
Yes I have enjoyed following your back-and-forths with John and have
not participated because I felt I had no useful information to add.
But I will pipe in when I think it useful or if I have questions of my
own.
About the "hijacking", I am responding to you separately.
John Harshman
Apr 10, 2013, 1:52:24 PM4/10/13
to
here.
John Harshman
Apr 10, 2013, 1:50:59 PM4/10/13
to
On 4/10/13 9:07 AM, pnyikos wrote:
> John, this thread has been pretty well hijacked in talk.origins by
> people making personal attacks, so I would like for us to move this on-
> topic discussion to sci.bio.paleontology. There, the majority of
> posts to this thread are still completely on-topic.
>
> Accordingly, I've added s.b.p. to the newsgroups and will be making
> on-topic comments after the attributions and earlier comments.
Sure. But all you will achieve with this is to fill s.b.p. with
> John, this thread has been pretty well hijacked in talk.origins by
> people making personal attacks, so I would like for us to move this on-
> topic discussion to sci.bio.paleontology. There, the majority of
> posts to this thread are still completely on-topic.
>
> Accordingly, I've added s.b.p. to the newsgroups and will be making
> on-topic comments after the attributions and earlier comments.
off-topic posts. Better if you just ignored the off-topic bits. I
realize that you are constitutionally unable to do so, but perhaps you
could at least try.
> I'll see if I can interest Richard Norman in joining us over there.
> His knowledge of paleontology could be of help.
>
> On Apr 4, 10:50 pm, John Harshman<jharsh...@pacbell.net> wrote:
>> On 4/4/13 11:54 AM, pnyikos wrote:
>
>>> On Mar 29, 12:47 pm, John Harshman<jharsh...@pacbell.net> wrote:
>>>> On 3/29/13 9:26 AM, pnyikos wrote:
>>
>>>>> On Mar 29, 7:04 am, Ron O<rokim...@cox.net> wrote:
>>>>>> Another interesting genome has been published. It comes from the same
>>>>>> lineage as crocs, dinos and birds.
>>
>>>>> This has been disputed by Gauthier and others, based on fossil
>>>>> evidence. They even dispute the claim that turtles are secondarily,
>>>>> rather than primarily, anapsid. If they are right, turtles are the
>>>>> outliers of extant Sauropsida.
>>
>>>> If. The molecular data have been getting more and more conclusive in the
>>>> past few years. Turtles are either archosaurs or the living sister group
>>>> of archosaurs.
>
> Still, Gauthier et. al. were holding out last year, according to the
> article whose abstract Ron O linked:
>
> 21. Lyson TR, Sperling EA, Heimberg AM, Gauthier JA, King BL, Peterson
> KJ:
> MicroRNAs support a turtle plus lizard clade. Biol Lett 2012,
> 8:104-107.
>
> -- http://genomebiology.com/content/pdf/gb-2013-14-3-r28.pdf
> See also a comment about two total analyses from their webpage [posted
> below]. But back to your last comment...
>
> Is there actually molecular support for turtles being more closely
> related to birds (in the cladistic sense) than to crocs??? That's
> what it would take for molecular data to show that turtles are
> archosaurs.
two hypotheses.
> At the opposite extreme, as long as only fragments of the genome of
> *Sphenodon* have been sequenced, it is too early to tell whether
> turtles are the living sister group of archosaurs, even from a
> molecular viewpoint.
of archosaurs. Do you have any basis for such a belief? If not, why is
Sphenodon relevant?
>>>> There is of course some morphological evidence that they are at least
>>>> diapsids, mostly advanced by Olivier Rieppel and colleagues.
>>
>>> Not here, unless one of the co-authors here is a colleague:
>>
>>>>>> http://genomebiology.com/2013/14/3/R28/abstract
>>
>>> Are you referring to the papers listed in the following passage?
>>
>>> "The position of turtles (Testudines) is uncertain; some authors place
>>> them approximately in the position shown above (Laurin& Reisz, 1995;
>>> Lee, 1993, 1995, 2001; Frost et al., 2006; Werneburg& S nchez-
>>> Villagra, 2009; Lyson et al. 2010), while others place them among
>>> Diapsida (deBraga& Rieppel, 1996, 1997; Rieppel& Reisz, 1999; Hedges
>>> & Poling, 1999; Mannen& Li, 1999; Hugall et al., 2007; Li et al.,
>>> 2008)."
>>> -- caption to phylogenetic tree of Amniota in
>>> http://tolweb.org/Amniota/14990
>>
>> The ones by Rieppel are among the ones I was referring to.
>
> Do you have a morphological analysis more recent than 1999 for turtles
> being diapsids? Gauthier et. al. have more recent ones to point to,
> including total analyses (see below).
>>> The tree shows turtles (Testudines) in Anapsida.
>>
>> Since the various references conflict, they picked one over the others.
>
> Seems they picked several 21st century ones over 20th century ones,
> where morphology is concerned.
conclusive.
>>> Laurin and Gauthier, who did the site, see your Rieppel et.al. ...1999
>>> and raise it to 2010.
>>
>>> Later down in the page they write:
>>
>>> "At least two total evidence analyses (in which molecular and
>>> morphological data are combined) suggest that turtles are not diapsids
>>> (Lee, 2001; Frost et al., 2006). In both, the molecular characters are
>>> much more numerous than the morphological ones, which implies
>>> significant support for the placement of turtles outside diapsids in
>>> these molecular datasets."
>
> Note the reasoning about the molecular evidence.
lots of uninformative characters. Or some combination. At any rate, the
molecular data have continued to expand at high speed, to the point
where any combined analysis will merely echo the molecular tree.
>>> I wonder who is cherry-picking this time, you or they.
>>
>> Neither.
>
> Why did you pick references from the 20th century, then? Was Rieppel
> just the first author that popped into your mind?
followed what he's done with it recently.
Richard Norman
Apr 10, 2013, 1:55:06 PM4/10/13
to
Quite frankly, you have not done the nearest thing which you can. You
should simply quit all that crap even if it gives you great
discomfort. You have the ability to dredge up all sorts of old
grudges and grievances and the ability to scrape open old wounds. You
also demonstrate that ability at enormous length and high frequency.
Please just give it up.
It doesn't matter if you have actually been aggrieved or whether you
simply perceive yourself aggrieved or whether you are just paranoid
and are making it all up. If people call you liar or worse and insult
you, then just let it lie. Those posts reflect more on their authors
and not on you. Your responses, even responses in kind and responses
you truly believe well justified , all those reflect on you and not to
your benefit.
Please do NOT respond to me by citing all the instances of when you
were insulted and "took the high road" only to be rebuffed or
whatever. Please especially do NOT tell me about all the terrible
things Ron O or others may have done to you and how you responded.
Please DO just let it drop. Bite your tongue, have a stiff drink,
yell at your dog -- whatever. But just keep your fingers off the
keyboard.
You do show interest in participating in real discussions, whether
about vertebrate evolution or directed panspermia. I may strongly
disagree with you about some of these issues and think you are way off
base. But once the discussion deviates from the subject matter and
turns to the personality and motivation of the people posting then it
is time to simply quit. Nobody will think "the other guy won because
you can't respond." People will only think "that God that's stopped."
jillery
Apr 10, 2013, 3:29:46 PM4/10/13
to
should be discussed.
John Harshman
Apr 10, 2013, 4:28:29 PM4/10/13
to
no ownership in usenet. Nobody decides, or perhaps everybody decides. I
think Peter was on-topic there, since he was talking about turtle
relationships, something that was also discussed in the original post.
You may think it isn't on-topic, but I would like to know your reasoning.
Richard Norman
Apr 10, 2013, 4:56:29 PM4/10/13
to
On Wed, 10 Apr 2013 13:28:29 -0700, John Harshman
Ron O did start the thread on the turtle genome and named it just that
in the subject line. Peter responded in kind about the turtle and had
a discussion quite on topic with John. Ron O immediately responded to
Peter's post with an attack on Peter. Can you hijack your own thread?
I think so.
I would prefer to return to the topic and so I ask John why
"traditional systematists" are held in such low regard. Isn't this
just a typical arrogance of the molecular people when there is
disagreement about lineage? Is there a consensus about turtles or is
there really some debate?
in the subject line. Peter responded in kind about the turtle and had
a discussion quite on topic with John. Ron O immediately responded to
Peter's post with an attack on Peter. Can you hijack your own thread?
I think so.
I would prefer to return to the topic and so I ask John why
"traditional systematists" are held in such low regard. Isn't this
just a typical arrogance of the molecular people when there is
disagreement about lineage? Is there a consensus about turtles or is
there really some debate?
John Harshman
Apr 10, 2013, 5:14:10 PM4/10/13
to
systematists but to those who prefer grades to clades, e.g. Ernst Mayr
and his (mostly also dead) compatriots. As for the arrogance of the
molecular people, it's by this point pretty well deserved. There's a lot
more data in the genome than elsewhere, and a lot more data in extant
species than in fossils.
> Is there a consensus about turtles or is
> there really some debate?
Then again it took a good 35 years (and counting) to get a consensus on
a matter so simple as the relationships of gharials, despite the data.
pnyikos
Apr 10, 2013, 6:30:30 PM4/10/13
to nyi...@bellsouth.net
Taking a short breather from the grading of tests...
On Apr 10, 5:46�pm, John Harshman <jharsh...@pacbell.net> wrote:
> On 4/10/13 2:34 PM, alias Ernest Major wrote:
> > On 10/04/2013 21:56, Richard Norman wrote:
> >> I would prefer to return to the topic and so I ask John why
> >> "traditional systematists" are held in such low regard. Isn't this
> >> just a typical arrogance of the molecular people when there is
> >> disagreement about lineage? Is there a consensus about turtles or is
> >> there really some debate?
>
> > I think you'll find that it's not morphological vs molecular, but
> > traditional vs cladistic. (I'm all that sure of the low regard either.)
I think it is molecular vs. morphological that takes into account
extinct taxa, like *Eusthenopteron*.
> > The problem with traditional systematics is baseing classifications on a
> > small number of subjectively selected "key" characters, rather than on
> > all the evidence.
Largely true, but subjectivity is in the eye of the beholder.
For instance, I've said in the past that just the structure of the
shoulder girdle in the monotremes vs. marsupials and placentals should
outweigh a really hefty number of minor (especially molecular)
characters as far as what is the sister group of what.
The scapula of marsupials and placentals is essentially identical,
even to a raised median ridge, while that of monotremes is utterly
different. Also there is at least one extra element in the monotreme
shoulder girdle.
Harshman's reply was a dismissive "If only it were that simple". Had
I said "all" except "a really hefty" [actually, words to that effect]
he would have been on target. Instead I think it would have been more
appropriate for him to say "I'm glad the process we use is not so
complicated."
[snip]
> That wasn't quite the "traditional" that Peter was talking about. He was
> referring, essentially, to gradism.
If you are referring to grades as in "subholostean grade" and
"holostean grade" you are dead wrong. That is the only example of
polyphyletic taxon that Romer gave in _Vertebrate Paleontology_, the
example of both the bowfin and the gar being in Holostei.
What I keep referring to is the traditional Linnean classification
whereby every organism had its own genus, family, order etc.
[including some intermediates like "suborder," "infraorder,"
"superfamily" in some cases]. Cladistic classification is great for
extant taxa, but the further back in time one goes, the more
unsatisfactory it is.
In guides to mushrooms, trees, etc. one has means of identification
for non-specialists that classifies things according to spore print
and other visible characters. If a fossil has the various apomorpies
that narrow down its clade missing, at least the Linnean
classification gave other ready criteria for identifying it, or at
least narrowing it down to the smallest feasible Linnean taxon.
> > And on the other hand one can criticise equal weighted maximum parsimony
> > as a technique - it leads to papers claiming that flight (re-)evolved n
> > times in stick insects, rather than being lost n+1 times.
I'm unfamilar with that example, but see above about radically
different weightings. Another example is megabats v. microbats. A
look at how the two inner digits of the wings are close together in
both, as well as the whole overall shape of the wings, should have
made everyone treat the hypothesis that they evolved flight separately
with the formula, "Extraordinary claims require extraordinary
evidence."
Ditto the claim that marsupials are sister group to
{placentals ,monotremes}.
> Oh, now, be fair. They had a likelihood model that gave them the same
> result, though I never quite understood why.
>
> > How to weight
> > traits is a difficult question to answer, but weighting them all the
> > same is clearly at best a simplifying approximation.
At worst, it is pseudoscience.
> > And selecting the
> > traits to use in an analysis remains a subjective process, leading
> > cherry picking still being a hazard.
I confine myself to clear cut cases like the two I've given above and
in the past.
Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer--
University of South Carolina
http://www.math.sc.edu/~nyikos/
> And, as in your stick insect example, the assumption of symmetrical
> transition probabilities may be unwarranted too.
On Apr 10, 5:46�pm, John Harshman <jharsh...@pacbell.net> wrote:
> On 4/10/13 2:34 PM, alias Ernest Major wrote:
> > On 10/04/2013 21:56, Richard Norman wrote:
> >> I would prefer to return to the topic and so I ask John why
> >> "traditional systematists" are held in such low regard. Isn't this
> >> just a typical arrogance of the molecular people when there is
> >> disagreement about lineage? Is there a consensus about turtles or is
> >> there really some debate?
>
> > traditional vs cladistic. (I'm all that sure of the low regard either.)
I think it is molecular vs. morphological that takes into account
extinct taxa, like *Eusthenopteron*.
> > The problem with traditional systematics is baseing classifications on a
> > small number of subjectively selected "key" characters, rather than on
> > all the evidence.
Largely true, but subjectivity is in the eye of the beholder.
For instance, I've said in the past that just the structure of the
shoulder girdle in the monotremes vs. marsupials and placentals should
outweigh a really hefty number of minor (especially molecular)
characters as far as what is the sister group of what.
The scapula of marsupials and placentals is essentially identical,
even to a raised median ridge, while that of monotremes is utterly
different. Also there is at least one extra element in the monotreme
shoulder girdle.
Harshman's reply was a dismissive "If only it were that simple". Had
I said "all" except "a really hefty" [actually, words to that effect]
he would have been on target. Instead I think it would have been more
appropriate for him to say "I'm glad the process we use is not so
complicated."
[snip]
> That wasn't quite the "traditional" that Peter was talking about. He was
> referring, essentially, to gradism.
If you are referring to grades as in "subholostean grade" and
"holostean grade" you are dead wrong. That is the only example of
polyphyletic taxon that Romer gave in _Vertebrate Paleontology_, the
example of both the bowfin and the gar being in Holostei.
What I keep referring to is the traditional Linnean classification
whereby every organism had its own genus, family, order etc.
[including some intermediates like "suborder," "infraorder,"
"superfamily" in some cases]. Cladistic classification is great for
extant taxa, but the further back in time one goes, the more
unsatisfactory it is.
In guides to mushrooms, trees, etc. one has means of identification
for non-specialists that classifies things according to spore print
and other visible characters. If a fossil has the various apomorpies
that narrow down its clade missing, at least the Linnean
classification gave other ready criteria for identifying it, or at
least narrowing it down to the smallest feasible Linnean taxon.
> > And on the other hand one can criticise equal weighted maximum parsimony
> > as a technique - it leads to papers claiming that flight (re-)evolved n
> > times in stick insects, rather than being lost n+1 times.
I'm unfamilar with that example, but see above about radically
different weightings. Another example is megabats v. microbats. A
look at how the two inner digits of the wings are close together in
both, as well as the whole overall shape of the wings, should have
made everyone treat the hypothesis that they evolved flight separately
with the formula, "Extraordinary claims require extraordinary
evidence."
Ditto the claim that marsupials are sister group to
{placentals ,monotremes}.
> Oh, now, be fair. They had a likelihood model that gave them the same
> result, though I never quite understood why.
>
> > How to weight
> > traits is a difficult question to answer, but weighting them all the
> > same is clearly at best a simplifying approximation.
At worst, it is pseudoscience.
> > And selecting the
> > traits to use in an analysis remains a subjective process, leading
> > cherry picking still being a hazard.
I confine myself to clear cut cases like the two I've given above and
in the past.
Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer--
University of South Carolina
http://www.math.sc.edu/~nyikos/
> transition probabilities may be unwarranted too.
pnyikos
Apr 10, 2013, 6:37:38 PM4/10/13
to nyi...@bellsouth.net
On Apr 10, 6:30�pm, pnyikos <nyik...@bellsouth.net> wrote:
> Taking a short breather from the grading of tests...
>
> On Apr 10, 5:46 pm, John Harshman <jharsh...@pacbell.net> wrote:
>
> > On 4/10/13 2:34 PM, alias Ernest Major wrote:
> > > I think you'll find that it's not morphological vs molecular, but
> > > traditional vs cladistic. (I'm all that sure of the low regard either.)
>
> I think it is molecular vs. morphological that takes into account
> extinct taxa, like *Eusthenopteron*.
Ouch! I meant *Eunotosaurus*.
> Taking a short breather from the grading of tests...
>
> On Apr 10, 5:46 pm, John Harshman <jharsh...@pacbell.net> wrote:
>
> > On 4/10/13 2:34 PM, alias Ernest Major wrote:
> > > I think you'll find that it's not morphological vs molecular, but
> > > traditional vs cladistic. (I'm all that sure of the low regard either.)
>
> I think it is molecular vs. morphological that takes into account
> extinct taxa, like *Eusthenopteron*.
It's been a long day already. :-(
John Harshman
Apr 10, 2013, 6:51:43 PM4/10/13
to
On 4/10/13 3:30 PM, pnyikos wrote:
> Taking a short breather from the grading of tests...
>
> On Apr 10, 5:46 pm, John Harshman<jharsh...@pacbell.net> wrote:
>> On 4/10/13 2:34 PM, alias Ernest Major wrote:
>
>>> On 10/04/2013 21:56, Richard Norman wrote:
>
>
>>>> I would prefer to return to the topic and so I ask John why
>>>> "traditional systematists" are held in such low regard. Isn't this
>>>> just a typical arrogance of the molecular people when there is
>>>> disagreement about lineage? Is there a consensus about turtles or is
>>>> there really some debate?
>>
>>> I think you'll find that it's not morphological vs molecular, but
>>> traditional vs cladistic. (I'm all that sure of the low regard either.)
>
> I think it is molecular vs. morphological that takes into account
> extinct taxa, like *Eusthenopteron*.
Then you are wrong. I think you're confusing the discussion of turtle
> Taking a short breather from the grading of tests...
>
> On Apr 10, 5:46 pm, John Harshman<jharsh...@pacbell.net> wrote:
>> On 4/10/13 2:34 PM, alias Ernest Major wrote:
>
>>> On 10/04/2013 21:56, Richard Norman wrote:
>
>
>>>> I would prefer to return to the topic and so I ask John why
>>>> "traditional systematists" are held in such low regard. Isn't this
>>>> just a typical arrogance of the molecular people when there is
>>>> disagreement about lineage? Is there a consensus about turtles or is
>>>> there really some debate?
>>
>>> I think you'll find that it's not morphological vs molecular, but
>>> traditional vs cladistic. (I'm all that sure of the low regard either.)
>
> I think it is molecular vs. morphological that takes into account
> extinct taxa, like *Eusthenopteron*.
cladistic relationships (not traditional systematics) with your ideas of
close relationships to ancestors (traditional).
>>> The problem with traditional systematics is baseing classifications on a
>>> small number of subjectively selected "key" characters, rather than on
>>> all the evidence.
>
> Largely true, but subjectivity is in the eye of the beholder.
>
> For instance, I've said in the past that just the structure of the
> shoulder girdle in the monotremes vs. marsupials and placentals should
> outweigh a really hefty number of minor (especially molecular)
> characters as far as what is the sister group of what.
>
> The scapula of marsupials and placentals is essentially identical,
> even to a raised median ridge, while that of monotremes is utterly
> different. Also there is at least one extra element in the monotreme
> shoulder girdle.
>
> Harshman's reply was a dismissive "If only it were that simple". Had
> I said "all" except "a really hefty" [actually, words to that effect]
> he would have been on target. Instead I think it would have been more
> appropriate for him to say "I'm glad the process we use is not so
> complicated."
to third parties.
>> That wasn't quite the "traditional" that Peter was talking about. He was
>> referring, essentially, to gradism.
>
> If you are referring to grades as in "subholostean grade" and
> "holostean grade" you are dead wrong. That is the only example of
> polyphyletic taxon that Romer gave in _Vertebrate Paleontology_, the
> example of both the bowfin and the gar being in Holostei.
>
> What I keep referring to is the traditional Linnean classification
> whereby every organism had its own genus, family, order etc.
> [including some intermediates like "suborder," "infraorder,"
> "superfamily" in some cases]. Cladistic classification is great for
> extant taxa, but the further back in time one goes, the more
> unsatisfactory it is.
It's possible to have a cladistic, Linnean classification. And of course
you like the traditional classification because it's all about grades.
> In guides to mushrooms, trees, etc. one has means of identification
> for non-specialists that classifies things according to spore print
> and other visible characters. If a fossil has the various apomorpies
> that narrow down its clade missing, at least the Linnean
> classification gave other ready criteria for identifying it, or at
> least narrowing it down to the smallest feasible Linnean taxon.
>>> And on the other hand one can criticise equal weighted maximum parsimony
>>> as a technique - it leads to papers claiming that flight (re-)evolved n
>>> times in stick insects, rather than being lost n+1 times.
>
> I'm unfamilar with that example, but see above about radically
> different weightings. Another example is megabats v. microbats. A
> look at how the two inner digits of the wings are close together in
> both, as well as the whole overall shape of the wings, should have
> made everyone treat the hypothesis that they evolved flight separately
> with the formula, "Extraordinary claims require extraordinary
> evidence."
>
> Ditto the claim that marsupials are sister group to
> {placentals ,monotremes}.
>
>> Oh, now, be fair. They had a likelihood model that gave them the same
>> result, though I never quite understood why.
>>
>>> How to weight
>>> traits is a difficult question to answer, but weighting them all the
>>> same is clearly at best a simplifying approximation.
>
> At worst, it is pseudoscience.
>>> And selecting the
>>> traits to use in an analysis remains a subjective process, leading
>>> cherry picking still being a hazard.
>
> I confine myself to clear cut cases like the two I've given above and
> in the past.
Richard Norman
Apr 10, 2013, 7:47:50 PM4/10/13
to
should have more weight than somebody elses. You say you have reasons
but then so does everybody else. In other words, it really isn't that
simple.
You may confine yourself to clear cut cases (whether the two you gave
are such is another problem) but isn't that a form of cherry picking?
You simply abandon anything tricky?
And "traditional" classification is more than Linnean categories. I
think a good example of grades is given by the quite traditional
notion of acoelomate, pseudocoelomate, and coelomate kinds of
protostome. Or non-vascular vs. vascular plants. But then I am not a
systemacist so I am not in a position to say what specialists in this
area consider "traditional" just as I am not in a position to evaluate
why this particular feature should count with heavier weight than that
one. I do know that counting as separate features those that are
really highly correlated by one developmental process is flawed.
jillery
Apr 10, 2013, 9:53:24 PM4/10/13
to
On Wed, 10 Apr 2013 13:28:29 -0700, John Harshman
If there is no ownership, then nobody has the authority to move the
topic to a different newsgroup.
Since the topic is "turtle genome sequence", the topic is explicitly
*not* "this thread has been pretty well hijacked in talk.origins by
people making personal attacks." So how you can say his comments
above are on-topic is beyond me.
You made substantially the same point here:
<rNKdnSEEZLS...@giganews.com>
so why you challenge me about it makes no sense. It's as if you think
only you have the authority to criticize him.
topic to a different newsgroup.
Since the topic is "turtle genome sequence", the topic is explicitly
*not* "this thread has been pretty well hijacked in talk.origins by
people making personal attacks." So how you can say his comments
above are on-topic is beyond me.
You made substantially the same point here:
<rNKdnSEEZLS...@giganews.com>
so why you challenge me about it makes no sense. It's as if you think
only you have the authority to criticize him.
Richard Norman
Apr 10, 2013, 10:16:58 PM4/10/13
to
On Wed, 10 Apr 2013 21:53:24 -0400, jillery <69jp...@gmail.com>
So where do you stand on the group once called "Anapsida"? Do holes
in the skull count for anything anymore?
in the skull count for anything anymore?
Glenn
Apr 10, 2013, 10:23:32 PM4/10/13
to
"jillery" <69jp...@gmail.com> wrote in message news:ij5cm8d14emn4gcq4...@4ax.com...
John Harshman
Apr 11, 2013, 1:07:28 AM4/11/13
to
> Since the topic is "turtle genome sequence", the topic is explicitly
> *not* "this thread has been pretty well hijacked in talk.origins by
> people making personal attacks." So how you can say his comments
> above are on-topic is beyond me.
> You made substantially the same point here:
>
> <rNKdnSEEZLS...@giganews.com>
>
> so why you challenge me about it makes no sense. It's as if you think
> only you have the authority to criticize him.
jillery
Apr 11, 2013, 2:03:13 AM4/11/13
to
On Wed, 10 Apr 2013 19:16:58 -0700, Richard Norman
<r_s_n...@comcast.net> wrote:
[...]
>So where do you stand on the group once called "Anapsida"? Do holes
>in the skull count for anything anymore?
I am no expert, but an interested observer, apparently like you. I
understand that in the past they had to rely on morphological
differences. It's reasonable to expect temporal fenestrae to be a
reliable indicator of lineage.
But I also recognize that morphological interpretation can be
subjective and not detailed enough to distinguish many cases. All
other things being equal, I would bet on the genetic data. So a
question in my mind is what evolutionary process drove turtles to
revert to the primitive form? Perhaps it had something to do with the
development of their shells, but that's just speculation.
<r_s_n...@comcast.net> wrote:
[...]
>So where do you stand on the group once called "Anapsida"? Do holes
>in the skull count for anything anymore?
understand that in the past they had to rely on morphological
differences. It's reasonable to expect temporal fenestrae to be a
reliable indicator of lineage.
But I also recognize that morphological interpretation can be
subjective and not detailed enough to distinguish many cases. All
other things being equal, I would bet on the genetic data. So a
question in my mind is what evolutionary process drove turtles to
revert to the primitive form? Perhaps it had something to do with the
development of their shells, but that's just speculation.
jillery
Apr 11, 2013, 11:31:13 AM4/11/13
to
On Wed, 10 Apr 2013 22:07:28 -0700, John Harshman
You conflate power with authority. Since you deny it, then how about
simple courtesy and a common sense application of the golden rule? One
could easily start a parallel topic in a different newsgroup, so
there's no need to move an exisiting one.
>> Since the topic is "turtle genome sequence", the topic is explicitly
>> *not* "this thread has been pretty well hijacked in talk.origins by
>> people making personal attacks." So how you can say his comments
>> above are on-topic is beyond me.
>
>I was referring to the on-topic comments that were also in the post.
In your post to which I replied, you make no such reference.
>> You made substantially the same point here:
>>
>> <rNKdnSEEZLS...@giganews.com>
>>
>> so why you challenge me about it makes no sense. It's as if you think
>> only you have the authority to criticize him.
>
>No authority is needed. But in fact most of his post was on-topic.
So you contradict yourself. No surprise there.
simple courtesy and a common sense application of the golden rule? One
could easily start a parallel topic in a different newsgroup, so
there's no need to move an exisiting one.
>> Since the topic is "turtle genome sequence", the topic is explicitly
>> *not* "this thread has been pretty well hijacked in talk.origins by
>> people making personal attacks." So how you can say his comments
>> above are on-topic is beyond me.
>
>I was referring to the on-topic comments that were also in the post.
>> You made substantially the same point here:
>>
>> <rNKdnSEEZLS...@giganews.com>
>>
>> so why you challenge me about it makes no sense. It's as if you think
>> only you have the authority to criticize him.
>
>No authority is needed. But in fact most of his post was on-topic.
jillery
Apr 11, 2013, 3:13:16 PM4/11/13
to
On Thu, 11 Apr 2013 10:43:45 -0700, John Harshman
<jhar...@pacbell.net> wrote:
>In fact, nobody has moved anything, if you will notice. Peter merely
>added s.b.p. to the newsgroups list. This is something worth arguing about?
You continue to criticize me for doing that which you do yourself at
the same time. That's a very silly strategy.
Re-read his post. He specifically wrote "so I would like for us to
move this on-topic discussion to sci.bio.paleontology." That you
removed that newsgroup from the distribution list puts the lie to your
claim that nobody is moving anything.
>>>> Since the topic is "turtle genome sequence", the topic is explicitly
>>>> *not* "this thread has been pretty well hijacked in talk.origins by
>>>> people making personal attacks." So how you can say his comments
>>>> above are on-topic is beyond me.
>>>
>>> I was referring to the on-topic comments that were also in the post.
>>
>> In your post to which I replied, you make no such reference.
>
>Didn't think an explicit reference was necessary, since you were
>presumably capable of reading the parts of Peters post you deleted.
The parts I deleted were not relevant to my point. You replied to my
post, not his. There was nothing of his post relevant to the topic in
my post. Your self-justifying illogic knows no bounds.
>>>> You made substantially the same point here:
>>>>
>>>> <rNKdnSEEZLS...@giganews.com>
>>>>
>>>> so why you challenge me about it makes no sense. It's as if you think
>>>> only you have the authority to criticize him.
>>>
>>> No authority is needed. But in fact most of his post was on-topic.
>>
>> So you contradict yourself. No surprise there.
>
>You have an odd tendency to find what isn't there. Different posts.
You have an odd tendency to ignore what is directly in front of you.
Different posts, same problem.
>The
>one you're responding to was largely on-topic.
Perhaps for variations of "largely" devoid of meaning.
>The one I responded to
>wasn't. Peter has a tendency to wander off into interminable, useless,
>off-topic arguments, but he doesn't always do that. And in this he is
>not alone.
Yeppers, you share that trait with him.
<jhar...@pacbell.net> wrote:
>added s.b.p. to the newsgroups list. This is something worth arguing about?
You continue to criticize me for doing that which you do yourself at
the same time. That's a very silly strategy.
Re-read his post. He specifically wrote "so I would like for us to
move this on-topic discussion to sci.bio.paleontology." That you
removed that newsgroup from the distribution list puts the lie to your
claim that nobody is moving anything.
>>>> Since the topic is "turtle genome sequence", the topic is explicitly
>>>> *not* "this thread has been pretty well hijacked in talk.origins by
>>>> people making personal attacks." So how you can say his comments
>>>> above are on-topic is beyond me.
>>>
>>> I was referring to the on-topic comments that were also in the post.
>>
>> In your post to which I replied, you make no such reference.
>
>presumably capable of reading the parts of Peters post you deleted.
The parts I deleted were not relevant to my point. You replied to my
post, not his. There was nothing of his post relevant to the topic in
my post. Your self-justifying illogic knows no bounds.
>>>> You made substantially the same point here:
>>>>
>>>> <rNKdnSEEZLS...@giganews.com>
>>>>
>>>> so why you challenge me about it makes no sense. It's as if you think
>>>> only you have the authority to criticize him.
>>>
>>> No authority is needed. But in fact most of his post was on-topic.
>>
>> So you contradict yourself. No surprise there.
>
You have an odd tendency to ignore what is directly in front of you.
Different posts, same problem.
>The
>one you're responding to was largely on-topic.
Perhaps for variations of "largely" devoid of meaning.
>The one I responded to
>wasn't. Peter has a tendency to wander off into interminable, useless,
>off-topic arguments, but he doesn't always do that. And in this he is
>not alone.
Yeppers, you share that trait with him.
Richard Norman
Apr 11, 2013, 6:38:36 PM4/11/13
to
On Thu, 11 Apr 2013 02:03:13 -0400, jillery <69jp...@gmail.com>
wrote:
Sorry for the delay but I have had some other business to attend to.
I am not sure what you mean by "revert to the primitive form". The
paper discusses a number of specialized turtle features, especially
longevity, resistance to anoxia resistance to cold, tooth loss, and
temperature determination of sex but I don't think any of these are
primitive. They also describe a rather slow mutation rate in turtles
compared with other amniotes so, if anything, there is "retention of
primitive form" rather than "reversion".
I recall turtles being a real problem for a very long time now.
Fortunately I never had to teach details of vertebrate evolution. For
intro biology it was enough to get across the notion of how different
amniote development was from fish-amphibian and how placental
development was just a slight modification of the general amniote egg.
There was a total of two hours (OK, 2 * 50 minutes) for this whole
business -- one on amniotes and one on special adaptations of birds
and mammals. So you just omit all the gory details -- just blindly
proceed ahead and, if one or two rare students ask questions about it
then you can discuss it privately. Just explaining why "reptiles" no
longer exist was a mess and the bird-dinosaur business just took three
words to describe: "Birds are dinosaurs. Next question?"
wrote:
I am not sure what you mean by "revert to the primitive form". The
paper discusses a number of specialized turtle features, especially
longevity, resistance to anoxia resistance to cold, tooth loss, and
temperature determination of sex but I don't think any of these are
primitive. They also describe a rather slow mutation rate in turtles
compared with other amniotes so, if anything, there is "retention of
primitive form" rather than "reversion".
I recall turtles being a real problem for a very long time now.
Fortunately I never had to teach details of vertebrate evolution. For
intro biology it was enough to get across the notion of how different
amniote development was from fish-amphibian and how placental
development was just a slight modification of the general amniote egg.
There was a total of two hours (OK, 2 * 50 minutes) for this whole
business -- one on amniotes and one on special adaptations of birds
and mammals. So you just omit all the gory details -- just blindly
proceed ahead and, if one or two rare students ask questions about it
then you can discuss it privately. Just explaining why "reptiles" no
longer exist was a mess and the bird-dinosaur business just took three
words to describe: "Birds are dinosaurs. Next question?"
John Harshman
Apr 11, 2013, 6:55:08 PM4/11/13
to
On 4/11/13 3:38 PM, Richard Norman wrote:
> On Thu, 11 Apr 2013 02:03:13 -0400, jillery<69jp...@gmail.com>
> wrote:
>
>> On Wed, 10 Apr 2013 19:16:58 -0700, Richard Norman
>> <r_s_n...@comcast.net> wrote:
>>
>> [...]
>>
>>> So where do you stand on the group once called "Anapsida"? Do holes
>>> in the skull count for anything anymore?
>>
>>
>> I am no expert, but an interested observer, apparently like you. I
>> understand that in the past they had to rely on morphological
>> differences. It's reasonable to expect temporal fenestrae to be a
>> reliable indicator of lineage.
>>
>> But I also recognize that morphological interpretation can be
>> subjective and not detailed enough to distinguish many cases. All
>> other things being equal, I would bet on the genetic data. So a
>> question in my mind is what evolutionary process drove turtles to
>> revert to the primitive form? Perhaps it had something to do with the
>> development of their shells, but that's just speculation.
>
> Sorry for the delay but I have had some other business to attend to.
>
> I am not sure what you mean by "revert to the primitive form".
I think she's talking about the amniote skull. Ancestral amniotes had no
> On Thu, 11 Apr 2013 02:03:13 -0400, jillery<69jp...@gmail.com>
> wrote:
>
>> On Wed, 10 Apr 2013 19:16:58 -0700, Richard Norman
>> <r_s_n...@comcast.net> wrote:
>>
>> [...]
>>
>>> So where do you stand on the group once called "Anapsida"? Do holes
>>> in the skull count for anything anymore?
>>
>>
>> I am no expert, but an interested observer, apparently like you. I
>> understand that in the past they had to rely on morphological
>> differences. It's reasonable to expect temporal fenestrae to be a
>> reliable indicator of lineage.
>>
>> But I also recognize that morphological interpretation can be
>> subjective and not detailed enough to distinguish many cases. All
>> other things being equal, I would bet on the genetic data. So a
>> question in my mind is what evolutionary process drove turtles to
>> revert to the primitive form? Perhaps it had something to do with the
>> development of their shells, but that's just speculation.
>
> Sorry for the delay but I have had some other business to attend to.
>
> I am not sure what you mean by "revert to the primitive form".
temporal fenestrae, ancestral diapsids had two, and turtles are back to
none. Hey, you're the one who brought up holes in the head.
Richard Norman
Apr 11, 2013, 7:29:04 PM4/11/13
to
It is just that I never bothered with the anatomy -- the physiology
was my thing and turtles had enough of that. The turtle heart is a
well known feature: it takes a licking and keeps on ticking. The
isolated heart will continue to beat for well over a day if you take
any reasonable care with it.
jillery
Apr 12, 2013, 4:11:47 AM4/12/13
to
On Thu, 11 Apr 2013 16:29:04 -0700, Richard Norman
IIUC my use of the term is standard terminology, I will explain, just
in case anybody else claims confusion.
"Primitive" in this context does not mean less sophisticated, but less
derived, closer to the basal form. For this topic, the basal form is
a solid skull, in contrast to a skull with temporal fenestrae.
The question of Anapsida paraphyly implies that species ancestral to
modern turtles were diapsids, and some turtle lineages lost their
temporal fenestrae over time. In its turn, that implication raises a
question; what evolutionary pressure caused these lineages to lose
what appears to be an advantageous adaptation?
The reasons for my question, and my question, are clear and obvious,
and my question is relevant to the topic.
in case anybody else claims confusion.
"Primitive" in this context does not mean less sophisticated, but less
derived, closer to the basal form. For this topic, the basal form is
a solid skull, in contrast to a skull with temporal fenestrae.
The question of Anapsida paraphyly implies that species ancestral to
modern turtles were diapsids, and some turtle lineages lost their
temporal fenestrae over time. In its turn, that implication raises a
question; what evolutionary pressure caused these lineages to lose
what appears to be an advantageous adaptation?
The reasons for my question, and my question, are clear and obvious,
and my question is relevant to the topic.
John Harshman
Apr 12, 2013, 10:53:10 AM4/12/13
to
as it's defined by a primitive condition, regardless of turtle
relationships.
> In its turn, that implication raises a
> question; what evolutionary pressure caused these lineages to lose
> what appears to be an advantageous adaptation?
fenestrae in the first place. Lightening the skull? Better distribution
of stress? More room for jaw muscle attachment? More room for jaw
muscles to bulge out? Turtles have another strategy for the last two:
excavating the rear of the skull.
> The reasons for my question, and my question, are clear and obvious,
> and my question is relevant to the topic.
Richard Norman
Apr 12, 2013, 11:07:30 AM4/12/13
to
On Fri, 12 Apr 2013 04:11:47 -0400, jillery <69jp...@gmail.com>
Your question may have been clear and obvious to most but I missed it
so I asked what you meant. John gave a simple and clear explanation,
one which I should have understood, so I apologized. And yes, all this
is smack on topic. If anyone else is lurking, here is a page about
holes in the head, stuff that I really did once know but apparently
forgot:
http://tolweb.org/notes/?note_id=463
so I asked what you meant. John gave a simple and clear explanation,
one which I should have understood, so I apologized. And yes, all this
is smack on topic. If anyone else is lurking, here is a page about
holes in the head, stuff that I really did once know but apparently
forgot:
http://tolweb.org/notes/?note_id=463
pnyikos
Apr 12, 2013, 9:58:39 PM4/12/13
to nyi...@bellsouth.net
On Apr 10, 6:51�ｿｽpm, John Harshman <jharsh...@pacbell.net> wrote:
> On 4/10/13 3:30 PM, pnyikos wrote:
> > On Apr 10, 5:46 pm, John Harshman<jharsh...@pacbell.net> �ｿｽwrote:
> On 4/10/13 3:30 PM, pnyikos wrote:
> >> On 4/10/13 2:34 PM, alias Ernest Major wrote:
>
> >>> On 10/04/2013 21:56, Richard Norman wrote:
>
> >>>> I would prefer to return to the topic and so I ask John why
> >>>> "traditional systematists" are held in such low regard. Isn't this
> >>>> just a typical arrogance of the molecular people when there is
> >>>> disagreement about lineage? Is there a consensus about turtles or is
> >>>> there really some debate?
>
> >>> I think you'll find that it's not morphological vs molecular, but
> >>> traditional vs cladistic. (I'm all that sure of the low regard either.)
>
> > I think it is molecular vs. morphological that takes into account
> > extinct taxa, like *Eusthenopteron*.
>
> Then you are wrong. I think you're confusing the discussion of turtle
> cladistic relationships (not traditional systematics) with your ideas of
> close relationships to ancestors (traditional).
You misunderstood my intent, and maybe I misunderstood Ernest Major's
>
> >>> On 10/04/2013 21:56, Richard Norman wrote:
>
> >>>> I would prefer to return to the topic and so I ask John why
> >>>> "traditional systematists" are held in such low regard. Isn't this
> >>>> just a typical arrogance of the molecular people when there is
> >>>> disagreement about lineage? Is there a consensus about turtles or is
> >>>> there really some debate?
>
> >>> I think you'll find that it's not morphological vs molecular, but
> >>> traditional vs cladistic. (I'm all that sure of the low regard either.)
>
> > I think it is molecular vs. morphological that takes into account
> > extinct taxa, like *Eusthenopteron*.
>
> Then you are wrong. I think you're confusing the discussion of turtle
> cladistic relationships (not traditional systematics) with your ideas of
> close relationships to ancestors (traditional).
intent. I was thinking of the last question Richard Norman asked, as
to wheret the controversy about turtle ancestry is centered.
A number of months ago I read an article by Gauthier in which he
described his latest morphological cladistic analysis, in which
extinct as well as extant species were incuded. Among them was the
enigmatic *Eunotosaurus* as well as the earliest known Chelonian. And
the result was that turtles were classed as anapsids.
> >>> The problem with traditional systematics is baseing classifications on a
> >>> small number of subjectively selected "key" characters, rather than on
> >>> all the evidence.
>
> > Largely true, but subjectivity is in the eye of the beholder.
>
> > For instance, I've said in the past that just the structure of the
> > shoulder girdle in the monotremes vs. marsupials and placentals should
> > outweigh a really hefty number of minor (especially molecular)
> > characters as far as what is the sister group of what.
>
> > The scapula of marsupials and placentals is essentially identical,
> > even to a raised median ridge, while that of monotremes is utterly
> > different.
series of fine drawings in Romer's _Vertebrate Paleontology showing
shoulder girdles of *Ophiacodon* (a pelycosaur), *Kannemeyeria* (a
dicynodont), the platypus, and the Virginia opossum. The first three
resemble each other far more than the last two resemble each other,
even when the extra elements of the first three shoulder girdles are
disregarded.
> >�ｿｽAlso there is at least one extra element in the monotreme
> > shoulder girdle.
Two. The interclavicle and the procoracoid.
> > Harshman's reply was a dismissive "If only it were that simple". �ｿｽHad
> > I said "all" except "a really hefty" [actually, words to that effect]
> > he would have been on target. �ｿｽInstead I think it would have been more
> > appropriate for him to say "I'm glad the process we use is not so
> > complicated."
>
> We'll all do better if you stop these retrospective judgments rendered
> to third parties.
You have complete freedom to update your opinion, but you do not avail
> > complicated."
>
> We'll all do better if you stop these retrospective judgments rendered
> to third parties.
yourself of it.
> >> That wasn't quite the "traditional" that Peter was talking about. He was
> >> referring, essentially, to gradism.
>
> > If you are referring to grades as in "subholostean grade" and
> > polyphyletic taxon that Romer gave in _Vertebrate Paleontology_, the
> > example of both the bowfin and the gar being in Holostei.
No reply to this from you, so I am still confused by what you mean by
> > example of both the bowfin and the gar being in Holostei.
"gradism." For now I will assume you really meant to refer to what I
described next:
> > What I keep referring to is the traditional Linnean classification
> > whereby every organism had its own genus, family, order etc.
> > [including some intermediates like "suborder," "infraorder,"
> > extant taxa, but the further back in time one goes, the more
> > unsatisfactory it is.
>
> You have confused cladistic classification with unranked classification.
> It's possible to have a cladistic, Linnean classification.
For extant animals, it is. But for extinct species it gets worse and
> > unsatisfactory it is.
>
> You have confused cladistic classification with unranked classification.
> It's possible to have a cladistic, Linnean classification.
worse the further back one goes. Organisms are all put at leaves due
to the taboo against depicting them in the position of ancestors. As
a result, stem species could have nothing besides a gigantic clade to
put them in, perhaps one containing the whole of tetrapoda, without
any family or even class to belong to.
In the traditional classification, these are in the paraphyletic class
Amphibia, and whether they are in a family or order all by themselves
is determined by how different they are from the nearest species in
morphology.
> And of course
> you like the traditional classification because it's all about grades.
because it is willing to pay any price for "objectivity," even at the
cost of being unable to orient oneself without putting untold hours
into studying minutiae.
> > In guides to mushrooms, trees, etc. one has means of identification
> > for non-specialists that classifies things according to spore print
> > that narrow down its clade missing, at least the Linnean
> > classification gave other ready criteria for identifying it, or at
> > least narrowing it down to the smallest feasible Linnean taxon.
>
> Not really, no.
You have a knack for bland denials. I hope what I say next won't get
> > classification gave other ready criteria for identifying it, or at
> > least narrowing it down to the smallest feasible Linnean taxon.
>
> Not really, no.
the same bland treatment.
Scene: a fossil hunter finds a scapula prartly buried in the ground
and asks an expert anatomist what he makes of it.
"Without removing more of the surrounding matrix, I can't say very
much except that it is a tetrapod, but neither a bird nor a marsupial
nor a placental. Do you know the age of this outcrop?"
"No, it's different from the ones we've identified so far, and the
geologist who could answer that question is already on his way home."
But after the better part of an hour, the fossil hunter uncovers
another bone from the skeleton and calls it to the attention of the
anatomist.
"That settles it!" the anatomist cries. "It's a mammal, perhaps a
monotreme. You've got a very significant find here, perhaps shedding
a lot of light on mammalian relationships. In the few days left to us
before we head for home, we should make every effort to unearth this
entire fossil to take with us. A find like this comes along only a
few times in a lifetime."
Concluded in next reply.
Peter Nyikos
Richard Norman
Apr 12, 2013, 10:32:53 PM4/12/13
to
opinions of some particular systemacists ignoring the fact that there
has been enormous controversy over just how to place turtles with an
enormity of disagreement. Romer's book, of course, is now almost 50
years old. Paleontologists, of course, generally have only morphology
of fossils to go on but that is not the be-all and end-all of
evolutionary information.
In particular, the post that started this thread cited an analysis of
the genome of a turtle. That paper concluded through a number of
independent lines of genetic information that turtles agreed more
closely with crocodilians and birds than with snakes and lizards. That
is not a pattern that you would expect for the anapsida.
pnyikos
Apr 12, 2013, 10:39:56 PM4/12/13
to nyi...@bellsouth.net
On Apr 10, 6:51�pm, John Harshman <jharsh...@pacbell.net> wrote:
> On 4/10/13 3:30 PM, pnyikos wrote:
> > On Apr 10, 5:46 pm, John Harshman<jharsh...@pacbell.net> �wrote:
> On 4/10/13 3:30 PM, pnyikos wrote:
> > On Apr 10, 5:46 pm, John Harshman<jharsh...@pacbell.net> �wrote:
> >>> How to weight
> >>> traits is a difficult question to answer, but weighting them all the
> >>> same is clearly at best a simplifying approximation.
>
> > At worst, it is pseudoscience.
>
> Why?
Because you might be giving as much weight to twenty, or whatever,
> >>> traits is a difficult question to answer, but weighting them all the
> >>> same is clearly at best a simplifying approximation.
>
> > At worst, it is pseudoscience.
>
> Why?
base pairs as you do to all the differences you have formally
described in the monotreme and placental-marsupial scapulas. Or to a
hundred base pairs as to all the similarities you see in the wings of
megabats and microbats.
And, as you know, although all base pairs are equal, some are far more
equal than others, resulting in huge differences in anatomy and mode
of life.
> >>> And selecting the
> >>> traits to use in an analysis remains a subjective process, leading
> >>> cherry picking still being a hazard.
>
> > I confine myself to clear cut cases like the two I've given above and
> > in the past.
>
> If only it were that simple, to coin a phrase.
bother with cases that are not so clear cut. But I will insist that
"extraordinary claims require extraordinary evidence" to anyone who
claims the two kinds of bats evolved flight independently, or that
monotremes form a clade with placentals that excludes marsupials.
Peter Nyikos
pnyikos
Apr 12, 2013, 10:43:36 PM4/12/13
to nyi...@bellsouth.net
On Apr 10, 7:47�pm, Richard Norman <r_s_nor...@comcast.net> wrote:
> On Wed, 10 Apr 2013 15:30:30 -0700 (PDT), pnyikos
> <nyik...@bellsouth.net> wrote:
> >On Apr 10, 5:46�pm, John Harshman <jharsh...@pacbell.net> wrote:
> >> On 4/10/13 2:34 PM, alias Ernest Major wrote:
> >> > And on the other hand one can criticise equal weighted maximum parsimony
> >> > as a technique - it leads to papers claiming that flight (re-)evolved n
> >> > times in stick insects, rather than being lost n+1 times.
>
> >I'm unfamilar with that example, but see above about radically
> >different weightings. �Another example is megabats v. microbats. �A
> >look at how the two inner digits of the wings are close together in
> >both, as well as the whole overall shape of the wings, should have
> >made everyone treat the hypothesis that they evolved flight separately
> >with the formula, "Extraordinary claims require extraordinary
> >evidence."
>
> >Ditto the claim that marsupials are sister group to
> >{placentals ,monotremes}.
[snip]
> On Wed, 10 Apr 2013 15:30:30 -0700 (PDT), pnyikos
> <nyik...@bellsouth.net> wrote:
> >On Apr 10, 5:46�pm, John Harshman <jharsh...@pacbell.net> wrote:
> >> On 4/10/13 2:34 PM, alias Ernest Major wrote:
> >> > And on the other hand one can criticise equal weighted maximum parsimony
> >> > as a technique - it leads to papers claiming that flight (re-)evolved n
> >> > times in stick insects, rather than being lost n+1 times.
>
> >I'm unfamilar with that example, but see above about radically
> >different weightings. �Another example is megabats v. microbats. �A
> >look at how the two inner digits of the wings are close together in
> >both, as well as the whole overall shape of the wings, should have
> >made everyone treat the hypothesis that they evolved flight separately
> >with the formula, "Extraordinary claims require extraordinary
> >evidence."
>
> >Ditto the claim that marsupials are sister group to
> >{placentals ,monotremes}.
> >> > And selecting the
> >> > traits to use in an analysis remains a subjective process, leading
> >> > cherry picking still being a hazard.
>
> >I confine myself to clear cut cases like the two I've given above and
> >in the past.
made to John a bit earlier this evening.
> > For instance, I've said in the past that just the structure of the
> > shoulder girdle in the monotremes vs. marsupials and placentals should
> > outweigh a really hefty number of minor (especially molecular)
> > characters as far as what is the sister group of what.
>
> > The scapula of marsupials and placentals is essentially identical,
> > even to a raised median ridge, while that of monotremes is utterly
> > different.
And harking back to therapsids and even pelycosaurs. There is a
series of fine drawings in Romer's _Vertebrate Paleontology showing
shoulder girdles of *Ophiacodon* (a pelycosaur), *Kannemeyeria* (a
dicynodont), the platypus, and the Virginia opossum. The first three
resemble each other far, far more than the last two resemble each
series of fine drawings in Romer's _Vertebrate Paleontology showing
shoulder girdles of *Ophiacodon* (a pelycosaur), *Kannemeyeria* (a
dicynodont), the platypus, and the Virginia opossum. The first three
other, even when the extra elements of the first three shoulder
girdles are disregarded.
girdles are disregarded.
> > Also there is at least one extra element in the monotreme
> > shoulder girdle.
> > shoulder girdle.
Two. The interclavicle and the procoracoid.
> I do not understand why your selection of morphologial �characters
> should have more weight than somebody elses. �You say you have reasons
> but then so does everybody else. �In other words, it really isn't that
> simple.
Take a look at a good reproduction of just the scapulas alone, and
> should have more weight than somebody elses. �You say you have reasons
> but then so does everybody else. �In other words, it really isn't that
> simple.
you should be struck by the difference. How can even a hundred base
pairs compete with that?
If you can't find any, I'm sure you can find detailed pictures of
megabat and microbat wings, and see how the claim that they
independently evolved from non-flying mammals was truly an
extraordinary one.
> You may confine yourself to clear cut cases (whether the two you gave
> are such is another problem) but isn't that a form of cherry picking?
> You simply abandon anything tricky?
where there are no well-established key characters that we can treat
the way Supreme Court justices described pornography, "We know it when
we see it."
> And "traditional" classification is more than Linnean categories. � I
> think a good example of grades is given by the quite traditional
> notion of acoelomate, pseudocoelomate, and coelomate kinds of
> protostome. �Or non-vascular vs. vascular plants.
ones are a paraphyletic group. Perhaps they are too big a hunk of
Plantae to comfortably fit into the limited number of Linnean grades,
but the classification has tolerated ungraded categories in the past.
> �But then I am not a
> systemacist so I am not in a position to say what specialists in this
> area consider "traditional" just as I am not in a position to evaluate
> why this particular feature should count with heavier weight than that
> one. �I do know that counting as separate features those that are
> really highly correlated by one developmental process is flawed.
Yes, that's another reason I say that weighting characters equally is
> area consider "traditional" just as I am not in a position to evaluate
> why this particular feature should count with heavier weight than that
> one. �I do know that counting as separate features those that are
> really highly correlated by one developmental process is flawed.
"at worst" pseudoscience.
Peter Nyikos
Richard Norman
Apr 12, 2013, 10:55:32 PM4/12/13
to
nucleotides, and the immediacy of the genetic data to the theoretical
foundations of evolutionary theory as opposed to the morphological
data being far down the line of phenotypic development gives
importance to the "new" thinking. There are a very large number of
cases where the genetic data has completely reorganized thinking about
evolutionary patterns that had been established for decades. The
"protostomes" have gone through massive revisions most recently just a
couple decades ago.
If you cherry pick the examples you are willing to discuss no doubt
you can find beautiful examples of particular morphological features
that "prove" your point. However if the genomic data contradicts
that, then which is to be believed? There are cases where all these
problems remain to be sorted out but I think a smart move is to put
your money on the molecular data.
pnyikos
Apr 12, 2013, 11:59:34 PM4/12/13
to nyi...@bellsouth.net
On Apr 12, 10:32�pm, Richard Norman <r_s_nor...@comcast.net> wrote:
> On Fri, 12 Apr 2013 18:58:39 -0700 (PDT), pnyikos
> <nyik...@bellsouth.net> wrote:
> On Fri, 12 Apr 2013 18:58:39 -0700 (PDT), pnyikos
> >On Apr 10, 6:51�pm, John Harshman <jharsh...@pacbell.net> wrote:
> >> On 4/10/13 3:30 PM, pnyikos wrote:
> >> > On Apr 10, 5:46 pm, John Harshman<jharsh...@pacbell.net> �wrote:
> >> On 4/10/13 3:30 PM, pnyikos wrote:
> >> >> On 4/10/13 2:34 PM, alias Ernest Major wrote:
>
> >> >>> On 10/04/2013 21:56, Richard Norman wrote:
>
> >> >>>> I would prefer to return to the topic and so I ask John why
> >> >>>> "traditional systematists" are held in such low regard. Isn't this
> >> >>>> just a typical arrogance of the molecular people when there is
> >> >>>> disagreement about lineage? Is there a consensus about turtles or is
> >> >>>> there really some debate?
>
> >> >>> I think you'll find that it's not morphological vs molecular, but
> >> >>> traditional vs cladistic. (I'm all that sure of the low regard either.)
>
> >> > I think it is molecular vs. morphological that takes into account
> >> > extinct taxa, like
*Eunotosaurus*.
>
> >> >>> On 10/04/2013 21:56, Richard Norman wrote:
>
> >> >>>> I would prefer to return to the topic and so I ask John why
> >> >>>> "traditional systematists" are held in such low regard. Isn't this
> >> >>>> just a typical arrogance of the molecular people when there is
> >> >>>> disagreement about lineage? Is there a consensus about turtles or is
> >> >>>> there really some debate?
>
> >> >>> I think you'll find that it's not morphological vs molecular, but
> >> >>> traditional vs cladistic. (I'm all that sure of the low regard either.)
>
> >> > I think it is molecular vs. morphological that takes into account
> >> > extinct taxa, like
>
> >> Then you are wrong. I think you're confusing the discussion of turtle
> >> cladistic relationships (not traditional systematics) with your ideas of
> >> close relationships to ancestors (traditional).
>
> >You misunderstood my intent, and maybe I misunderstood Ernest Major's
> >intent. �I was thinking of the last question Richard Norman asked, as
>
> >A number of months ago I read an article by Gauthier in which he
> >described his latest morphological cladistic analysis, in which
> >extinct as well as extant species were incuded. �Among them was the
> >enigmatic *Eunotosaurus* as well as the earliest known Chelonian. �And
> >the result was that turtles were classed as anapsids.
The above was followed (spatially, not chronologically) by a long
> >A number of months ago I read an article by Gauthier in which he
> >described his latest morphological cladistic analysis, in which
> >extinct as well as extant species were incuded. �Among them was the
> >enigmatic *Eunotosaurus* as well as the earliest known Chelonian. �And
> >the result was that turtles were classed as anapsids.
discussion on a completely different topic:
> >> >>> The problem with traditional systematics is baseing classifications on a
> >> >>> small number of subjectively selected "key" characters, rather than on
> >> >>> all the evidence.
> You seem fixated on some particular morphological features and the
> opinions of some particular systemacists ignoring the fact that there
> has been enormous controversy over just how to place turtles with an
> enormity of disagreement.
with turtles. Gauthier et.al. did not just focus on some particular
morphological features in that article to which I was referring. They
used a proper cladistic analysis.
I believe I was the first person in talk.origins to call attention to
this striking discrepancy between the placement of turtles by a
molecular study and their traditional classification as anapsids.
At the beginning of 1998, Laurence Moran posted about an article on
phylogeny as suggested by molecular analysis of alpha and beta
hemoglobin, and myoglobin. The authors briefly mentioned a feature
that they noticed was unusual: mammals being closer to birds than
either was to lizards and snakes.
However, they (and also Moran) had overlooked an anomaly I found
equally striking: turtles within archosauria, with birds being more
closely related (cladistically speaking) to them and to Sphenodon than
to crocodilians. Here is the post where I made a big fanfare out of
it--it seemed so radical at the time:
http://groups.google.com/group/talk.origins/msg/c98d0b20052ff803
And here is an excerpt from another post:
===================== begin excerpt
See the above paper, where the performance is mixed. Five incorrect
(IMO) groupings had bootstrap values over 50 for NJ. Removing the
viper led to a much more sensible arrangement--except that the
bootstrap NJ value for grouping non-lizards, including mammals,
away from lizards shot up to 80. With the viper included,
it was below 40. But then, how can you trust a system
which puts the viper as more closely related to horses and
humans than to ANY of the other reptiles, including lizards?
To be sure, the bootstrap values for putting the viper with
us were below 40%, but that was what was keeping
the bootstrap value for the {birds, crocodilians, mammals,
turtles, Spenondon} group away from lizards from shooting up to 80.
=========================== end of excerpt
>�Romer's book, of course, is now almost 50
> years old. �Paleontologists, of course, generally have only morphology
> of fossils to go on but that is not the be-all and end-all of
> evolutionary information.
I never said it was, and neither did Gauthier. Recall the words I
> of fossils to go on but that is not the be-all and end-all of
> evolutionary information.
quoted to Harshman:
"At least two total evidence analyses (in which molecular and
morphological data are combined) suggest that turtles are not
diapsids
(Lee, 2001; Frost et al., 2006). In both, the molecular characters
are
much more numerous than the morphological ones, which implies
significant support for the placement of turtles outside diapsids in
these molecular datasets."
I can't see why Harshman had difficulty with this passage. To me it
seems perfectly straightforward: if the molecular evidence were
overwhelmingly in favor of turtles being diapsids then their sheer
number would have caused these two analyses to put turtles into
Diapsida. But in fact they did not do that, so the molecular evidence
would seem to be only slightly in favor of them being diapsids.
> In particular, the post that started this thread cited an analysis of
> the genome of a turtle. �That paper concluded through a number of
> independent lines of genetic information that turtles agreed more
> closely with crocodilians and birds than with snakes and lizards.
outside the diapsid clade, or in a clade with snakes and lizards. If
you look at the diagrams in the huge linked pdf file, you will see
some of them.
> That
> is not a pattern that you would expect for the anapsida.
Peter Nyikos
alias Ernest Major
Apr 13, 2013, 7:38:16 AM4/13/13
to
glaucophytes, green algae and "land" plants). I have the impression that
the widespread acceptance of this as a clade has a lot to do with a
"morphological" trait (possession of primary plastids). However the
molecular data is so far far from clear cut on relationships among
non-excavate bikonts (Corticata).
Plastids appear to form a clade within cyanobacteria, from which it is
inferred that endosymbiosis occurred only once, and Archaeplastida are
monophyletic. Alternative hypotheses include
1) Endosymbiosis occurred several times involving a clade of
cyanobacteria no longer extant (or not yet sampled).
2) Endosymbiosis occurred several times involving several cyanobacterial
clades and plastid monophyly is only apparent. Convergence does occur in
plastid evolution - for example the same genes have been lost from the
plastome in several different plant clades. The question is whether it
is plausible that this could result in sufficient convergence at the
sequence level.
3) Some Archaeplastids (putatively some or all of Hacrobia) lost their
primary plastids, and subsquently gained secondary plastids.
--
alias Ernest Major
John Harshman
Apr 13, 2013, 9:29:14 AM4/13/13
to
your posts confusing and annoying: long quotes from yourself, years ago;
Replies to material that's several levels of attribution in the past;
extended off-topic asides; and so on. Dial it back.
> http://groups.google.com/group/talk.origins/msg/c98d0b20052ff803
>
> And here is an excerpt from another post:
>
> ===================== begin excerpt
>
> See the above paper, where the performance is mixed. Five incorrect
> (IMO) groupings had bootstrap values over 50 for NJ. Removing the
> viper led to a much more sensible arrangement--except that the
> bootstrap NJ value for grouping non-lizards, including mammals,
> away from lizards shot up to 80. With the viper included,
> it was below 40. But then, how can you trust a system
> which puts the viper as more closely related to horses and
> humans than to ANY of the other reptiles, including lizards?
>
> To be sure, the bootstrap values for putting the viper with
> us were below 40%, but that was what was keeping
> the bootstrap value for the {birds, crocodilians, mammals,
> turtles, Spenondon} group away from lizards from shooting up to 80.
>
> =========================== end of excerpt
>> Romer's book, of course, is now almost 50
>> years old. Paleontologists, of course, generally have only morphology
>> of fossils to go on but that is not the be-all and end-all of
>> evolutionary information.
>
> I never said it was, and neither did Gauthier. Recall the words I
> quoted to Harshman:
>
> "At least two total evidence analyses (in which molecular and
> morphological data are combined) suggest that turtles are not
> diapsids
> (Lee, 2001; Frost et al., 2006). In both, the molecular characters
> are
> much more numerous than the morphological ones, which implies
> significant support for the placement of turtles outside diapsids in
> these molecular datasets."
>
> I can't see why Harshman had difficulty with this passage.
> To me it
> seems perfectly straightforward: if the molecular evidence were
> overwhelmingly in favor of turtles being diapsids then their sheer
> number would have caused these two analyses to put turtles into
> Diapsida. But in fact they did not do that, so the molecular evidence
> would seem to be only slightly in favor of them being diapsids.
molecular evidence might fail to be decisive. The authors picked one. I
mentioned others. Let's try again. First, morphological data are usually
pruned of parsimony-uninformative characters, mostly invariant and
autapomorphic ones, before analysis. DNA sequence data are not. So the
simple size of the data set is deceptive, and the sequence data set is
effectively smaller than it looks. Second, there is often much noise,
i.e. characters some of which match all possible topologies. What the
authors allege is something else: a strong, secondary signal nearly
equal to the primary signal, causing conflict that renders the
conclusion vulnerable to small perturbations. Sometimes that does
happen, and there are ways to test for it. The quote doesn't say if the
authors tried any of those methods, but my point is that the evidence
they adduce in that quote isn't useful.
Here are a couple of papers that do attempt to assess secondary signal
in a valid way (though not in turtles): Gatesy, J., P. O'Grady, and R.
H. Baker. 1999. Corroboration among data sets in simultaneous analysis:
Hidden support for phylogenetic relationships among higher level
artiodactyl taxa. Cladistics 15:271-313; Harshman, J., C. J. Huddleston,
J. Bollback, T. M. Parsons, and M. J. Braun. 2003. True and false
gharials: A nuclear gene phylogeny of Crocodylia. Systematic Biology
52:386-402.
>> In particular, the post that started this thread cited an analysis of
>> the genome of a turtle. That paper concluded through a number of
>> independent lines of genetic information that turtles agreed more
>> closely with crocodilians and birds than with snakes and lizards.
>
> But there were also a number of lines where the outcome had them
> outside the diapsid clade, or in a clade with snakes and lizards. If
> you look at the diagrams in the huge linked pdf file, you will see
> some of them.
figure, fig. 5, which shows squamates sister to all other amniotes.The
other 5 genes in that figure show turtles either as sister to archosaurs
or in a trichotomy with birds and crocodiles.
John Harshman
Apr 13, 2013, 9:33:33 AM4/13/13
to
On 4/12/13 7:39 PM, pnyikos wrote:
> On Apr 10, 6:51 pm, John Harshman<jharsh...@pacbell.net> wrote:
>> On 4/10/13 3:30 PM, pnyikos wrote:
>
>>> On Apr 10, 5:46 pm, John Harshman<jharsh...@pacbell.net> wrote:
>>>>> How to weight
>>>>> traits is a difficult question to answer, but weighting them all the
>>>>> same is clearly at best a simplifying approximation.
>>
>>> At worst, it is pseudoscience.
>>
>> Why?
>
> Because you might be giving as much weight to twenty, or whatever,
> base pairs as you do to all the differences you have formally
> described in the monotreme and placental-marsupial scapulas. Or to a
> hundred base pairs as to all the similarities you see in the wings of
> megabats and microbats.
How do you decide how much weight to give?
> On Apr 10, 6:51 pm, John Harshman<jharsh...@pacbell.net> wrote:
>> On 4/10/13 3:30 PM, pnyikos wrote:
>
>>> On Apr 10, 5:46 pm, John Harshman<jharsh...@pacbell.net> wrote:
>>>>> How to weight
>>>>> traits is a difficult question to answer, but weighting them all the
>>>>> same is clearly at best a simplifying approximation.
>>
>>> At worst, it is pseudoscience.
>>
>> Why?
>
> Because you might be giving as much weight to twenty, or whatever,
> base pairs as you do to all the differences you have formally
> described in the monotreme and placental-marsupial scapulas. Or to a
> hundred base pairs as to all the similarities you see in the wings of
> megabats and microbats.
> And, as you know, although all base pairs are equal, some are far more
> equal than others, resulting in huge differences in anatomy and mode
> of life.
the ones you should pay less attention to, as they would be most subject
to convergence. Completely useless and trivial characters should, if
anything, get the strongest weight.
>>>>> And selecting the
>>>>> traits to use in an analysis remains a subjective process, leading
>>>>> cherry picking still being a hazard.
>>
>>> I confine myself to clear cut cases like the two I've given above and
>>> in the past.
>>
>> If only it were that simple, to coin a phrase.
>
> It is that simple, because I am not a professional and do not have to
> bother with cases that are not so clear cut. But I will insist that
> "extraordinary claims require extraordinary evidence" to anyone who
> claims the two kinds of bats evolved flight independently, or that
> monotremes form a clade with placentals that excludes marsupials.
otherwise. There is strong evidence from many sources, in each case. By
the way, there is a fine, magic bullet character linking turtles to
diapsids, if you're looking for one. They all have a hooked 5th
metatarsal. So does that convince you? I would hope not.
alias Ernest Major
Apr 13, 2013, 10:05:17 AM4/13/13
to
On 13/04/2013 14:33, John Harshman wrote:
>> And, as you know, although all base pairs are equal, some are far more
>> equal than others, resulting in huge differences in anatomy and mode
>> of life.
>
> It would seem to me that the ones that result in huge differences are
> the ones you should pay less attention to, as they would be most subject
> to convergence. Completely useless and trivial characters should, if
> anything, get the strongest weight.
For molecular traits, does it not depend on the time depth? Neutrally
>> And, as you know, although all base pairs are equal, some are far more
>> equal than others, resulting in huge differences in anatomy and mode
>> of life.
>
> It would seem to me that the ones that result in huge differences are
> the ones you should pay less attention to, as they would be most subject
> to convergence. Completely useless and trivial characters should, if
> anything, get the strongest weight.
evolving base pairs are good for resolving the topology of recent
clades, but if you go to a sufficiently old clade they're effectively
randomised, and noise overwhlems the signal.
--
alias Ernest Major
jillery
Apr 13, 2013, 12:20:23 PM4/13/13
to
about "none"?
Richard Norman
Apr 13, 2013, 12:44:11 PM4/13/13
to
triumphs of the past", real or perceived. Gautier's work dates back to
the late 80's and Rieppel's to the mid 90's. That you mentioned
problems in the late 90's about turtles may or may not have been the
first mention here on talk.origins but certainly was not new to
biology. The simple fact is that there has long been some dispute
over the placement of turtles which means there is some evidence to
support one position but different evidence to support a different
position.
*Hemidactylus*
Apr 13, 2013, 12:56:06 PM4/13/13
to
On 04/10/2013 12:36 PM, pnyikos wrote:
> CC: Richard Norman, because of the invitation below.
>
> On Apr 4, 8:36 pm, Richard Norman <r_s_nor...@comcast.net> wrote:
>> On Thu, 4 Apr 2013 16:52:30 -0700 (PDT), Ron O <rokim...@cox.net>
>> wrote:
> [huge snip]
>
>>> I'm the one that has had enough. I still can't understand why anyone
>>> would want to read junk that involves Nyikos.
>
> On this thread, it would presumably be the fact that I've kept on-
> topic on this thread, except for one reply to Ron O. That continued
> today with a reply I did to John Harshman a short while ago. It had
> to do with the evidence of where turtles sit in Sauropsida-- diapsid
> v. anapsid, archosaur v. sister group of archosaurs v. sister group of
> {archosaurs, tuatara} v. Anapsida.
>
> Richard, I was hoping you could join the two of us on this
> discussion. With your knowledge of paleontology, I think you can be a
> big help in discussing the morphological evidence, at least.
>
> I've added sci.bio.paleontology to the newsgroups because (see my
> reply to Harshman) this thread has been pretty well trashed by all the
> personal attacks by others, and there we have a good chance of keeping
> it on-topic. In fact, I wouldn't mind if you both started posting
> just to s.b.p. I'll be watching for you over there.
>
>>> I stopped reading most
>>> of the crap Nyikos posts 2 years ago.
>>
>> If you stopped reading it then how come you respond so vociferously
>> and at such enormous length? Why do you think we tolerate your crap
>> about him any better than his rants? Why don't both of you just give
>> it a rest? Even one of you calling it quits would be an enormous
>> improvement.
>
> I've done the nearest thing to that which I comfortably can. On the
> "By Their Fruits...." thread I've told Hemidactylus about my new
> policy. I gave him an almost complete statement of it, as follows:
>
> "from now on, I will confine myself to very
> brief direct replies to posts of Ron O,"
>
> --http://groups.google.com/group/talk.origins/msg/5ec655abe34712f9
>
> I carelessly left off a conditional bit ("in reply to me") that I had
> put in a much earlier post to Hemidactylus:
>
> "But after the third knockdown, I will reply to almost
> nothing Ron O posts in reply to me..."
>
> -- http://groups.google.com/group/talk.origins/msg/d319bac1c0c85623
>
> Note: "almost nothing," not "almost no posts". That could come
> later, if enough people become aware of my new policy. Everyone on
> this thread seems to be oblivious to it so far, despite my hewing to
> it in the one off-topic post I've done to this thread so far.
I think if you could declare a self-imposed moratorium on replying to
interpersonal stuff aimed at you for several weeks to a month and just
focus on turtle phylogeny, your DP FAQ and other non-personal stuff and
if others would self-impose the same with regard to you, the tensions
might subside. The rest of us would welcome the new direction if you and
others you've been battling with could agree to those terms.
> CC: Richard Norman, because of the invitation below.
>
> On Apr 4, 8:36 pm, Richard Norman <r_s_nor...@comcast.net> wrote:
>> On Thu, 4 Apr 2013 16:52:30 -0700 (PDT), Ron O <rokim...@cox.net>
>> wrote:
> [huge snip]
>
>>> I'm the one that has had enough. I still can't understand why anyone
>>> would want to read junk that involves Nyikos.
>
> On this thread, it would presumably be the fact that I've kept on-
> topic on this thread, except for one reply to Ron O. That continued
> today with a reply I did to John Harshman a short while ago. It had
> to do with the evidence of where turtles sit in Sauropsida-- diapsid
> v. anapsid, archosaur v. sister group of archosaurs v. sister group of
> {archosaurs, tuatara} v. Anapsida.
>
> Richard, I was hoping you could join the two of us on this
> discussion. With your knowledge of paleontology, I think you can be a
> big help in discussing the morphological evidence, at least.
>
> I've added sci.bio.paleontology to the newsgroups because (see my
> reply to Harshman) this thread has been pretty well trashed by all the
> personal attacks by others, and there we have a good chance of keeping
> it on-topic. In fact, I wouldn't mind if you both started posting
> just to s.b.p. I'll be watching for you over there.
>
>>> I stopped reading most
>>> of the crap Nyikos posts 2 years ago.
>>
>> If you stopped reading it then how come you respond so vociferously
>> and at such enormous length? Why do you think we tolerate your crap
>> about him any better than his rants? Why don't both of you just give
>> it a rest? Even one of you calling it quits would be an enormous
>> improvement.
>
> I've done the nearest thing to that which I comfortably can. On the
> "By Their Fruits...." thread I've told Hemidactylus about my new
> policy. I gave him an almost complete statement of it, as follows:
>
> "from now on, I will confine myself to very
> brief direct replies to posts of Ron O,"
>
> --http://groups.google.com/group/talk.origins/msg/5ec655abe34712f9
>
> I carelessly left off a conditional bit ("in reply to me") that I had
> put in a much earlier post to Hemidactylus:
>
> "But after the third knockdown, I will reply to almost
> nothing Ron O posts in reply to me..."
>
> -- http://groups.google.com/group/talk.origins/msg/d319bac1c0c85623
>
> Note: "almost nothing," not "almost no posts". That could come
> later, if enough people become aware of my new policy. Everyone on
> this thread seems to be oblivious to it so far, despite my hewing to
> it in the one off-topic post I've done to this thread so far.
I think if you could declare a self-imposed moratorium on replying to
interpersonal stuff aimed at you for several weeks to a month and just
focus on turtle phylogeny, your DP FAQ and other non-personal stuff and
if others would self-impose the same with regard to you, the tensions
might subside. The rest of us would welcome the new direction if you and
others you've been battling with could agree to those terms.
jillery
Apr 13, 2013, 1:40:23 PM4/13/13
to
read from anybody. Congratulations.
Of course, there are always truce violations, real, imagined, and
provoked. But let's see if he's willing to try. There's no sense
starting out cynical.
John Harshman
Apr 13, 2013, 4:28:39 PM4/13/13
to
posting.
John Harshman
Apr 13, 2013, 4:30:38 PM4/13/13
to
jillery
Apr 13, 2013, 4:52:08 PM4/13/13
to
On Sat, 13 Apr 2013 13:28:39 -0700, John Harshman
I replied directly to what you wrote. If my reply is not on-topic,
then neither is what you wrote. That you wrote other things which I
did not reply to does not alter that fact.
then neither is what you wrote. That you wrote other things which I
did not reply to does not alter that fact.
John Harshman
Apr 13, 2013, 8:43:53 PM4/13/13
to
On 4/12/13 6:58 PM, pnyikos wrote:
>>>>> The problem with traditional systematics is baseing classifications on a
>>>>> small number of subjectively selected "key" characters, rather than on
>>>>> all the evidence.
>>
>>> Largely true, but subjectivity is in the eye of the beholder.
>>
>>> For instance, I've said in the past that just the structure of the
>>> shoulder girdle in the monotremes vs. marsupials and placentals should
>>> outweigh a really hefty number of minor (especially molecular)
>>> characters as far as what is the sister group of what.
>>
>>> The scapula of marsupials and placentals is essentially identical,
>>> even to a raised median ridge, while that of monotremes is utterly
>>> different.
>
> And harking back to therapsids and even pelycosaurs. There is a
> series of fine drawings in Romer's _Vertebrate Paleontology showing
> shoulder girdles of *Ophiacodon* (a pelycosaur), *Kannemeyeria* (a
> dicynodont), the platypus, and the Virginia opossum. The first three
> resemble each other far more than the last two resemble each other,
> even when the extra elements of the first three shoulder girdles are
> disregarded.
You're responding to yourself two layers back again. Try to refrain.
>>> Also there is at least one extra element in the monotreme
>>> shoulder girdle.
>
> Two. The interclavicle and the procoracoid.
>
>>> Harshman's reply was a dismissive "If only it were that simple". Had
>>>> That wasn't quite the "traditional" that Peter was talking about. He was
>>>> referring, essentially, to gradism.
>>
>>> If you are referring to grades as in "subholostean grade" and
>>> "holostean grade" you are dead wrong. That is the only example of
"Thecodont" is a grade. "Mammal-like reptile" is a grade. Traditional
"Amphibia" is a grade. And so on. Traditional systematists, including
Simpson and Romer, liked grades.
>>> What I keep referring to is the traditional Linnean classification
>>> whereby every organism had its own genus, family, order etc.
>>> [including some intermediates like "suborder," "infraorder,"
>>> "superfamily" in some cases]. Cladistic classification is great for
data necessary to do so.
> As
> a result, stem species could have nothing besides a gigantic clade to
> put them in, perhaps one containing the whole of tetrapoda, without
> any family or even class to belong to.
If you know of any clear stem species, tell me how you know.
> In the traditional classification, these are in the paraphyletic class
> Amphibia, and whether they are in a family or order all by themselves
> is determined by how different they are from the nearest species in
> morphology.
In other words, gradism.
>> And of course
>> you like the traditional classification because it's all about grades.
>
> And all the advantages they provide. And you like the cladistic
> because it is willing to pay any price for "objectivity," even at the
> cost of being unable to orient oneself without putting untold hours
> into studying minutiae.
You find it easier to orient yourself with traditional grades purely
because they're familiar to you.
>>> In guides to mushrooms, trees, etc. one has means of identification
>>> for non-specialists that classifies things according to spore print
>>> and other visible characters. If a fossil has the various apomorpies
any idea where you're trying to go.
> On Apr 10, 6:51 pm, John Harshman<jharsh...@pacbell.net> wrote:
>> On 4/10/13 3:30 PM, pnyikos wrote:
>> On 4/10/13 3:30 PM, pnyikos wrote:
>>> On Apr 10, 5:46 pm, John Harshman<jharsh...@pacbell.net> wrote:
>>>> On 4/10/13 2:34 PM, alias Ernest Major wrote:
>>
>>>>> On 10/04/2013 21:56, Richard Norman wrote:
>>
>>>>>> I would prefer to return to the topic and so I ask John why
>>>>>> "traditional systematists" are held in such low regard. Isn't this
>>>>>> just a typical arrogance of the molecular people when there is
>>>>>> disagreement about lineage? Is there a consensus about turtles or is
>>>>>> there really some debate?
>>
>>>>> I think you'll find that it's not morphological vs molecular, but
>>>>> traditional vs cladistic. (I'm all that sure of the low regard either.)
>>
>>> I think it is molecular vs. morphological that takes into account
>>> extinct taxa, like *Eusthenopteron*.
>>
>> Then you are wrong. I think you're confusing the discussion of turtle
>> cladistic relationships (not traditional systematics) with your ideas of
>> close relationships to ancestors (traditional).
>
> You misunderstood my intent, and maybe I misunderstood Ernest Major's
> intent. I was thinking of the last question Richard Norman asked, as
> to wheret the controversy about turtle ancestry is centered.
>
> A number of months ago I read an article by Gauthier in which he
> described his latest morphological cladistic analysis, in which
> extinct as well as extant species were incuded. Among them was the
> enigmatic *Eunotosaurus* as well as the earliest known Chelonian. And
> the result was that turtles were classed as anapsids.
All true. But whatever does that have to do with traditional systematists?
>>>> On 4/10/13 2:34 PM, alias Ernest Major wrote:
>>
>>>>> On 10/04/2013 21:56, Richard Norman wrote:
>>
>>>>>> I would prefer to return to the topic and so I ask John why
>>>>>> "traditional systematists" are held in such low regard. Isn't this
>>>>>> just a typical arrogance of the molecular people when there is
>>>>>> disagreement about lineage? Is there a consensus about turtles or is
>>>>>> there really some debate?
>>
>>>>> I think you'll find that it's not morphological vs molecular, but
>>>>> traditional vs cladistic. (I'm all that sure of the low regard either.)
>>
>>> I think it is molecular vs. morphological that takes into account
>>> extinct taxa, like *Eusthenopteron*.
>>
>> Then you are wrong. I think you're confusing the discussion of turtle
>> cladistic relationships (not traditional systematics) with your ideas of
>> close relationships to ancestors (traditional).
>
> You misunderstood my intent, and maybe I misunderstood Ernest Major's
> intent. I was thinking of the last question Richard Norman asked, as
> to wheret the controversy about turtle ancestry is centered.
>
> A number of months ago I read an article by Gauthier in which he
> described his latest morphological cladistic analysis, in which
> extinct as well as extant species were incuded. Among them was the
> enigmatic *Eunotosaurus* as well as the earliest known Chelonian. And
> the result was that turtles were classed as anapsids.
>>>>> The problem with traditional systematics is baseing classifications on a
>>>>> small number of subjectively selected "key" characters, rather than on
>>>>> all the evidence.
>>
>>> Largely true, but subjectivity is in the eye of the beholder.
>>
>>> For instance, I've said in the past that just the structure of the
>>> shoulder girdle in the monotremes vs. marsupials and placentals should
>>> outweigh a really hefty number of minor (especially molecular)
>>> characters as far as what is the sister group of what.
>>
>>> The scapula of marsupials and placentals is essentially identical,
>>> even to a raised median ridge, while that of monotremes is utterly
>>> different.
>
> And harking back to therapsids and even pelycosaurs. There is a
> series of fine drawings in Romer's _Vertebrate Paleontology showing
> shoulder girdles of *Ophiacodon* (a pelycosaur), *Kannemeyeria* (a
> dicynodont), the platypus, and the Virginia opossum. The first three
> resemble each other far more than the last two resemble each other,
> even when the extra elements of the first three shoulder girdles are
> disregarded.
>>> Also there is at least one extra element in the monotreme
>>> shoulder girdle.
>
> Two. The interclavicle and the procoracoid.
>
>>> I said "all" except "a really hefty" [actually, words to that effect]
>>> he would have been on target. Instead I think it would have been more
>>> appropriate for him to say "I'm glad the process we use is not so
>>> complicated."
>>
>> We'll all do better if you stop these retrospective judgments rendered
>> to third parties.
>
> You have complete freedom to update your opinion, but you do not avail
> yourself of it.
I see no need. Instead, I prefer to discourage these digressions.
>>> complicated."
>>
>> We'll all do better if you stop these retrospective judgments rendered
>> to third parties.
>
> You have complete freedom to update your opinion, but you do not avail
> yourself of it.
>>>> That wasn't quite the "traditional" that Peter was talking about. He was
>>>> referring, essentially, to gradism.
>>
>>> If you are referring to grades as in "subholostean grade" and
>>> polyphyletic taxon that Romer gave in _Vertebrate Paleontology_, the
>>> example of both the bowfin and the gar being in Holostei.
>
> No reply to this from you, so I am still confused by what you mean by
> "gradism." For now I will assume you really meant to refer to what I
> described next:
Not at all. Most grades are paraphyletic. "Reptile" is a grade.
>>> example of both the bowfin and the gar being in Holostei.
>
> No reply to this from you, so I am still confused by what you mean by
> "gradism." For now I will assume you really meant to refer to what I
> described next:
"Thecodont" is a grade. "Mammal-like reptile" is a grade. Traditional
"Amphibia" is a grade. And so on. Traditional systematists, including
Simpson and Romer, liked grades.
>>> What I keep referring to is the traditional Linnean classification
>>> whereby every organism had its own genus, family, order etc.
>>> [including some intermediates like "suborder," "infraorder,"
>>> extant taxa, but the further back in time one goes, the more
>>> unsatisfactory it is.
>>
>> You have confused cladistic classification with unranked classification.
>> It's possible to have a cladistic, Linnean classification.
>
> For extant animals, it is. But for extinct species it gets worse and
> worse the further back one goes. Organisms are all put at leaves due
> to the taboo against depicting them in the position of ancestors.
This isn't a taboo. It's a practical recognition that we can't have the
>>> unsatisfactory it is.
>>
>> You have confused cladistic classification with unranked classification.
>> It's possible to have a cladistic, Linnean classification.
>
> For extant animals, it is. But for extinct species it gets worse and
> worse the further back one goes. Organisms are all put at leaves due
> to the taboo against depicting them in the position of ancestors.
data necessary to do so.
> As
> a result, stem species could have nothing besides a gigantic clade to
> put them in, perhaps one containing the whole of tetrapoda, without
> any family or even class to belong to.
> In the traditional classification, these are in the paraphyletic class
> Amphibia, and whether they are in a family or order all by themselves
> is determined by how different they are from the nearest species in
> morphology.
>> And of course
>> you like the traditional classification because it's all about grades.
>
> And all the advantages they provide. And you like the cladistic
> because it is willing to pay any price for "objectivity," even at the
> cost of being unable to orient oneself without putting untold hours
> into studying minutiae.
because they're familiar to you.
>>> In guides to mushrooms, trees, etc. one has means of identification
>>> for non-specialists that classifies things according to spore print
>>> that narrow down its clade missing, at least the Linnean
>>> classification gave other ready criteria for identifying it, or at
>>> least narrowing it down to the smallest feasible Linnean taxon.
>>
>> Not really, no.
>
> You have a knack for bland denials. I hope what I say next won't get
> the same bland treatment.
>
> Scene: a fossil hunter finds a scapula prartly buried in the ground
> and asks an expert anatomist what he makes of it.
>
> "Without removing more of the surrounding matrix, I can't say very
> much except that it is a tetrapod, but neither a bird nor a marsupial
> nor a placental. Do you know the age of this outcrop?"
>
> "No, it's different from the ones we've identified so far, and the
> geologist who could answer that question is already on his way home."
>
> But after the better part of an hour, the fossil hunter uncovers
> another bone from the skeleton and calls it to the attention of the
> anatomist.
>
> "That settles it!" the anatomist cries. "It's a mammal, perhaps a
> monotreme. You've got a very significant find here, perhaps shedding
> a lot of light on mammalian relationships. In the few days left to us
> before we head for home, we should make every effort to unearth this
> entire fossil to take with us. A find like this comes along only a
> few times in a lifetime."
>
> Concluded in next reply.
I will presumably respond in the next reply, because so far I don't have
>>> classification gave other ready criteria for identifying it, or at
>>> least narrowing it down to the smallest feasible Linnean taxon.
>>
>> Not really, no.
>
> You have a knack for bland denials. I hope what I say next won't get
> the same bland treatment.
>
> Scene: a fossil hunter finds a scapula prartly buried in the ground
> and asks an expert anatomist what he makes of it.
>
> "Without removing more of the surrounding matrix, I can't say very
> much except that it is a tetrapod, but neither a bird nor a marsupial
> nor a placental. Do you know the age of this outcrop?"
>
> "No, it's different from the ones we've identified so far, and the
> geologist who could answer that question is already on his way home."
>
> But after the better part of an hour, the fossil hunter uncovers
> another bone from the skeleton and calls it to the attention of the
> anatomist.
>
> "That settles it!" the anatomist cries. "It's a mammal, perhaps a
> monotreme. You've got a very significant find here, perhaps shedding
> a lot of light on mammalian relationships. In the few days left to us
> before we head for home, we should make every effort to unearth this
> entire fossil to take with us. A find like this comes along only a
> few times in a lifetime."
>
> Concluded in next reply.
any idea where you're trying to go.
jillery
Apr 15, 2013, 2:56:52 AM4/15/13
to
On Sat, 13 Apr 2013 17:34:10 -0700, John Harshman
<jhar...@pacbell.net> wrote:
>Sorry, I replied to you by mistake. I'll try not to in the future.
At least.
<jhar...@pacbell.net> wrote:
At least.
pnyikos
Apr 15, 2013, 12:26:47 PM4/15/13
to nyi...@bellsouth.net
On Apr 10, 1:50�pm, John Harshman <jharsh...@pacbell.net> wrote:
> On 4/10/13 9:07 AM, pnyikos wrote:
>
> > John, this thread has been pretty well hijacked in talk.origins by
> > people making personal attacks, so I would like for us to move this on-
> > topic discussion to sci.bio.paleontology. �There, the majority of
> > posts to this thread are still completely on-topic.
>
> > Accordingly, I've added s.b.p. to the newsgroups and will be making
> > on-topic comments after the attributions and earlier comments.
>
> Sure. But all you will achieve with this is to fill s.b.p. with
> off-topic posts.
Fortunately, that has not happened to a great extent.
> Better if you just ignored the off-topic bits. I
> realize that you are constitutionally unable to do so, but perhaps you
> could at least try.
As you can see, I've really given it the old college try so far; also,
I've kept personal attacks out of even the off-topic remarks I've
made.
But now, back to an on-topic matter of great interest to me:
[snip for focus]
> > At the opposite extreme, as long as only fragments of the genome of
> > *Sphenodon* have been sequenced, it is too early to tell whether
> > turtles are the living sister group of archosaurs, even from a
> > molecular viewpoint.
>
> Only if you think there's some chance that Sphenodon is the sister group
> of archosaurs. Do you have any basis for such a belief?
Yes, in the paper by Fushitani et. al. that I alluded to last week:
Fushitani K, Higashiyama K, Moriyama EN, Imai K, Hosokawa K,
Evolution of Amniote Hb alpha Globin Sequences
Mol Biol Evol 1996 Sep;13(7):1039-1043
The following cladogram is part of a bigger one in the paper:
______________Caiman
|
|
|100/98
____|
| |_____________Alligator
|
|
--------| 96/68
| ___________Turtle
| |
| | _______Ostrich
|____| |
|58/27 |
| ________|90/92
| | |
| | |________Chicken
|__ |
|74/32
|
|
|_____________Sphenodon
The first number is for NJ, the latter for maximum parsimony.
The sister group for the above was the pair of mammals,
rather than the lizards. With the viper left out of the analysis,
the NJ bootstrap value for this "sister grouping" was 80, but
maximum parsimony was below 50.
The viper really made a difference; more about that in a separate
post.
The moral is that, as long as we have only fragmentary information
about the Sphenodon genome, its relationships are up in the air.
Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer--
University of South Carolina
http://www.math.sc.edu/~nyikos/
nyikos @ math.sc.edu
> On 4/10/13 9:07 AM, pnyikos wrote:
>
> > John, this thread has been pretty well hijacked in talk.origins by
> > people making personal attacks, so I would like for us to move this on-
> > topic discussion to sci.bio.paleontology. �There, the majority of
> > posts to this thread are still completely on-topic.
>
> > Accordingly, I've added s.b.p. to the newsgroups and will be making
> > on-topic comments after the attributions and earlier comments.
>
> Sure. But all you will achieve with this is to fill s.b.p. with
> off-topic posts.
Fortunately, that has not happened to a great extent.
> Better if you just ignored the off-topic bits. I
> realize that you are constitutionally unable to do so, but perhaps you
> could at least try.
As you can see, I've really given it the old college try so far; also,
I've kept personal attacks out of even the off-topic remarks I've
made.
But now, back to an on-topic matter of great interest to me:
[snip for focus]
> > At the opposite extreme, as long as only fragments of the genome of
> > *Sphenodon* have been sequenced, it is too early to tell whether
> > turtles are the living sister group of archosaurs, even from a
> > molecular viewpoint.
>
> Only if you think there's some chance that Sphenodon is the sister group
> of archosaurs. Do you have any basis for such a belief?
Yes, in the paper by Fushitani et. al. that I alluded to last week:
Fushitani K, Higashiyama K, Moriyama EN, Imai K, Hosokawa K,
Evolution of Amniote Hb alpha Globin Sequences
Mol Biol Evol 1996 Sep;13(7):1039-1043
The following cladogram is part of a bigger one in the paper:
______________Caiman
|
|
|100/98
____|
| |_____________Alligator
|
|
--------| 96/68
| ___________Turtle
| |
| | _______Ostrich
|____| |
|58/27 |
| ________|90/92
| | |
| | |________Chicken
|__ |
|74/32
|
|
|_____________Sphenodon
The first number is for NJ, the latter for maximum parsimony.
The sister group for the above was the pair of mammals,
rather than the lizards. With the viper left out of the analysis,
the NJ bootstrap value for this "sister grouping" was 80, but
maximum parsimony was below 50.
The viper really made a difference; more about that in a separate
post.
The moral is that, as long as we have only fragmentary information
about the Sphenodon genome, its relationships are up in the air.
Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer--
University of South Carolina
http://www.math.sc.edu/~nyikos/
nyikos @ math.sc.edu
pnyikos
Apr 15, 2013, 12:59:01 PM4/15/13
to nyi...@bellsouth.net
> >http://groups.google.com/group/talk.origins/msg/c98d0b20052ff803
>
> > And here is an excerpt from another post:
>
> > ===================== begin excerpt
>
> > See the above paper, where the performance is mixed. �Five incorrect
> > (IMO) groupings had bootstrap values over 50 for NJ. �Removing the
> > viper led to a much more sensible arrangement--except that the
> > bootstrap NJ value for grouping non-lizards, including mammals,
> > away from lizards shot up to 80. �With the viper included,
> > it was below 40. �But then, how can you trust a system
> > which puts the viper as more closely related to horses and
> > humans than to ANY of the other reptiles, including lizards?
>
> > To be sure, the bootstrap values for putting the viper with
> > us were below 40%, but that was what was keeping
> > the bootstrap value for the {birds, crocodilians, mammals,
> > turtles, Spenondon} group away from lizards from shooting up to 80.
>
> > =========================== �end of excerpt
>
> All that, and you never actually say what your present point is. What is it?
I was hoping someone would comment on the obvious volatility of
> All that, and you never actually say what your present point is. What is it?
bootstrap values -- how including one more taxon can really affect
them, without me having to get explicit about them.
> >> � Romer's book, of course, is now almost 50
> >> years old. �Paleontologists, of course, generally have only morphology
> >> of fossils to go on but that is not the be-all and end-all of
> >> evolutionary information.
>
> > I never said it was, and neither did Gauthier. �Recall the words I
> > quoted to Harshman:
>
> > "At least two total evidence analyses (in which molecular and
> > morphological data are combined) suggest that turtles are not
> > diapsids
> > (Lee, 2001; Frost et al., 2006). In both, the molecular characters
> > are
> > much more numerous than the morphological ones, which implies
> > significant support for the placement of turtles outside diapsids in
> > these molecular datasets."
>
> > I can't see why Harshman had difficulty with this passage.
> Harshman didn't have difficulty. *You* had difficulty.
Looks to me like we both had difficulties, but of a different nature.
> >> of fossils to go on but that is not the be-all and end-all of
> >> evolutionary information.
>
> > I never said it was, and neither did Gauthier. �Recall the words I
> > quoted to Harshman:
>
> > "At least two total evidence analyses (in which molecular and
> > morphological data are combined) suggest that turtles are not
> > diapsids
> > (Lee, 2001; Frost et al., 2006). In both, the molecular characters
> > are
> > much more numerous than the morphological ones, which implies
> > significant support for the placement of turtles outside diapsids in
> > these molecular datasets."
>
> > I can't see why Harshman had difficulty with this passage.
> Harshman didn't have difficulty. *You* had difficulty.
> > To me it
> > seems perfectly straightforward: if the molecular evidence were
> > overwhelmingly in favor of turtles being diapsids then their sheer
> > number would have caused these two analyses to put turtles into
> > Diapsida. �But in fact they did not do that, so the molecular evidence
> > would seem to be only slightly in favor of them being diapsids.
>
> That isn't what the passage says.
is what you think it *ought to* have said.
> There are many possible reasons why
> molecular evidence might fail to be decisive. The authors picked one. I
> mentioned others. Let's try again. First, morphological data are usually
> pruned of parsimony-uninformative characters, mostly invariant and
> autapomorphic ones, before analysis. DNA sequence data are not. So the
> simple size of the data set is deceptive, and the sequence data set is
> effectively smaller than it looks.
analysis. Didn't you mean to say "effectively bigger than it looks?"
> Second, there is often much noise,
> i.e. characters some of which match all possible topologies. What the
> authors allege is something else: a strong, secondary signal nearly
> equal to the primary signal, causing conflict that renders the
> conclusion vulnerable to small perturbations.
groupings being so sensitive to the inclusion of one more taxon in the
passage I quoted from long ago?
pnyikos
Apr 15, 2013, 1:15:28 PM4/15/13
to nyi...@bellsouth.net
On Apr 13, 9:29�am, John Harshman <jharsh...@pacbell.net> wrote:
> On 4/12/13 8:59 PM, pnyikos wrote:
Has Giganews started putting times automatically into attribution
> On 4/12/13 8:59 PM, pnyikos wrote:
lines? If so, this one is four hours off from EDT and nine off from
GMT. I did the post to which you are following up just before
midnight, and it shows at the end.
Note how the post to which I was following up is given a later time by
Google.
[big snip of things covered in first reply, and a little more]
> >> In particular, the post that started this thread cited an analysis of
> >> the genome of a turtle. �That paper concluded through a number of
> >> independent lines of genetic information that turtles agreed more
> >> closely with crocodilians and birds than with snakes and lizards.
>
> > But there were also a number of lines where the outcome had them
> > outside the diapsid clade, or in a clade with snakes and lizards. �If
> > you look at the diagrams in the huge linked pdf file, you will see
> > some of them.
mammals are put in a clade with archosaurs, away from lizards, wrt
SOX9, and with lizards and away from archosaurs wrt two other genes,
DMRT1 and WT1.
IOW, I was confusing one of the big apparent anomalies (wrt alpha
hemoglobin, back then) that I've been aware of for over a decade and a
half with the other.
Moral: I should have called it quits well before midnight.
> What diagrams are you talking about? I found one gene, SOX9, in one
> figure, fig. 5, which shows squamates sister to all other amniotes.
Like in those older analyses of blood factors. I wonder what is
behind it. The split between Sauropsida and Synapsida is supposed to
go back to well before the Permian.
> The
> other 5 genes in that figure show turtles either as sister to archosaurs
> or in a trichotomy with birds and crocodiles.
Peter Nyikos
John Harshman
Apr 15, 2013, 9:57:43 PM4/15/13
to
some snake sequences are odd. Also, I would never use neighbor-joining,
since its performance can be weird under various circumstances. I'm
suspecting that there wasn't all that much sequence involved here, at
least by modern standards.
>>>> Romer's book, of course, is now almost 50
>>>> years old. Paleontologists, of course, generally have only morphology
>>>> of fossils to go on but that is not the be-all and end-all of
>>>> evolutionary information.
>>
>>> I never said it was, and neither did Gauthier. Recall the words I
>>> quoted to Harshman:
>>
>>> "At least two total evidence analyses (in which molecular and
>>> morphological data are combined) suggest that turtles are not
>>> diapsids
>>> (Lee, 2001; Frost et al., 2006). In both, the molecular characters
>>> are
>>> much more numerous than the morphological ones, which implies
>>> significant support for the placement of turtles outside diapsids in
>>> these molecular datasets."
>>
>>> I can't see why Harshman had difficulty with this passage.
>
>> Harshman didn't have difficulty. *You* had difficulty.
>
> Looks to me like we both had difficulties, but of a different nature.
>>> To me it
>>> seems perfectly straightforward: if the molecular evidence were
>>> overwhelmingly in favor of turtles being diapsids then their sheer
>>> number would have caused these two analyses to put turtles into
>>> Diapsida. But in fact they did not do that, so the molecular evidence
>>> would seem to be only slightly in favor of them being diapsids.
>>
>> That isn't what the passage says.
>
> It *is* what the passage literally *says*; what you talk about below
> is what you think it *ought to* have said.
>> There are many possible reasons why
>> molecular evidence might fail to be decisive. The authors picked one. I
>> mentioned others. Let's try again. First, morphological data are usually
>> pruned of parsimony-uninformative characters, mostly invariant and
>> autapomorphic ones, before analysis. DNA sequence data are not. So the
>> simple size of the data set is deceptive, and the sequence data set is
>> effectively smaller than it looks.
>
> Huh? When you prune characters, you reduce the number in the
> analysis. Didn't you mean to say "effectively bigger than it looks?"
morphological data a large proportion of the characters are useless.
>> Second, there is often much noise,
>> i.e. characters some of which match all possible topologies. What the
>> authors allege is something else: a strong, secondary signal nearly
>> equal to the primary signal, causing conflict that renders the
>> conclusion vulnerable to small perturbations.
>
> Interesting. Is this somehow related to the bootstrap values for those
> groupings being so sensitive to the inclusion of one more taxon in the
> passage I quoted from long ago?
that the use of neighbor-joining had something to do with it.
John Harshman
Apr 15, 2013, 10:02:32 PM4/15/13
to
On 4/15/13 10:15 AM, pnyikos wrote:
> On Apr 13, 9:29 am, John Harshman<jharsh...@pacbell.net> wrote:
>> On 4/12/13 8:59 PM, pnyikos wrote:
>
> Has Giganews started putting times automatically into attribution
> lines? If so, this one is four hours off from EDT and nine off from
> GMT. I did the post to which you are following up just before
> midnight, and it shows at the end.
I don't know.
> On Apr 13, 9:29 am, John Harshman<jharsh...@pacbell.net> wrote:
>> On 4/12/13 8:59 PM, pnyikos wrote:
>
> Has Giganews started putting times automatically into attribution
> lines? If so, this one is four hours off from EDT and nine off from
> GMT. I did the post to which you are following up just before
> midnight, and it shows at the end.
>>>> On Apr 12, 10:32 pm, Richard Norman<r_s_nor...@comcast.net> wrote:
>
> Note how the post to which I was following up is given a later time by
> Google.
>
> [big snip of things covered in first reply, and a little more]
>
>>>> In particular, the post that started this thread cited an analysis of
>>>> the genome of a turtle. That paper concluded through a number of
>>>> independent lines of genetic information that turtles agreed more
>>>> closely with crocodilians and birds than with snakes and lizards.
>>
>>> But there were also a number of lines where the outcome had them
>>> outside the diapsid clade, or in a clade with snakes and lizards. If
>>> you look at the diagrams in the huge linked pdf file, you will see
>>> some of them.
>
> My memory was playing tricks on me again. I was mis-remembering how
> mammals are put in a clade with archosaurs, away from lizards, wrt
> SOX9, and with lizards and away from archosaurs wrt two other genes,
> DMRT1 and WT1.
They show a polytomy for mammals, lepidosaurs, and archosauromorphs.
> IOW, I was confusing one of the big apparent anomalies (wrt alpha
> hemoglobin, back then) that I've been aware of for over a decade and a
> half with the other.
>
> Moral: I should have called it quits well before midnight.
>
>> What diagrams are you talking about? I found one gene, SOX9, in one
>> figure, fig. 5, which shows squamates sister to all other amniotes.
>
> Like in those older analyses of blood factors. I wonder what is
> behind it. The split between Sauropsida and Synapsida is supposed to
> go back to well before the Permian.
>> The
>> other 5 genes in that figure show turtles either as sister to archosaurs
>> or in a trichotomy with birds and crocodiles.
>
> Not quite. WT1 has turtles in a group with crocs, away from birds.
John Harshman
Apr 15, 2013, 10:07:40 PM4/15/13
to
DNA sequence isn't all that reliable.
pnyikos
Apr 16, 2013, 3:44:14 PM4/16/13
to nyi...@bellsouth.net
http://groups.google.com/group/sci.bio.systematics/msg/e96c7a47a6fca219
Message-ID: <2001061415...@kappa.math.sc.edu>
You actually replied to that post way back then, but you didn't
comment on the cladogram. Well, better late than never.
> > The viper really made a difference; more about that in a separate
> > post.
> > The moral is that, �as long as we have only fragmentary information
> > about the Sphenodon genome, its relationships are up in the air.
formulation seems more natural to me:
> No, the moral is that a neighbor-joining analysis of a short
> DNA sequence isn't all that reliable.
Please keep my conditional "as long as" in mind.
Peter Nyikos
John Harshman
Apr 16, 2013, 3:56:06 PM4/16/13
to
big enough that the lower reaches of that distance would still be quite
sufficient for practical certainty about relationships. That is, the
paper you reference had perhaps a few times 10^2 bases, the genome is a
few times 10^9, and a few times 10^3 or 10^4 would almost certainly do
the trick. So "fragmentary information" is plenty. Just not that
fragmentary. And NJ was probably an especially poor choice of analytic
methods.
pnyikos
Apr 16, 2013, 9:49:59 PM4/16/13
to nyi...@bellsouth.net
>
> > You actually replied to that post way back then, but you didn't
> > comment on the cladogram. �Well, better late than never.
>
> >>> The viper really made a difference; more about that in a separate
> >>> post.
The wording of the following sentence is alluded to below.
> > You actually replied to that post way back then, but you didn't
> > comment on the cladogram. �Well, better late than never.
>
> >>> The viper really made a difference; more about that in a separate
> >>> post.
> >>> The moral is that, �as long as we have only fragmentary information
> >>> about the Sphenodon genome, its relationships are up in the air.
> >> No, the moral is that a neighbor-joining analysis of a short
> >> DNA sequence isn't all that reliable.
>
> > Do you have any reason whatsoever for thinking *my* moral is false?
> > Please keep my conditional "as long as" in mind.
>
> Yes. The distance between a small, single gene and an entire genome is
> big enough that the lower reaches of that distance would still be quite
> sufficient for practical certainty about relationships. That is, the
> paper you reference had perhaps a few times 10^2 bases, the genome is a
> few times 10^9, and a few times 10^3 or 10^4 would almost certainly do
> the trick.
on these things?] but I did not quantify "only fragmentary" and I had
in mind the current state of fragmentariness of what we know of the
Sphenodon genome.
Right near the beginning of this thread, I said that the omission of
Sphenodon was a serious flaw in the paper we started out discussing.
A little later, the following exchange ensued between us:
_________excerpt, you going first and last_______________
>> There is no apparent correlation, based on all sorts of data in
all
>> sorts of taxa, between morphological and molecular evolutionary rates.
> However, I am still curious to know what the situation is with
> Sphenodon, as suggested in my first reply.
You must remain curious until there's a Sphenodon genome or a
substantial portion thereof.
============ end of excerpt===========
> So "fragmentary information" is plenty. Just not that
> fragmentary.
vocabulary. I take it that not even a measly 10^4 base pairs of
Sphenodon (the tuatara) have been sequenced in the 17 years since
those blood proteins were sequenced.
In New Zealand between then and now (2001, if memory serves) I saw
some live tuataras in captivity. Surely the sample needed to get lots
of base pairs would have been small and harmless.
> And NJ was probably an especially poor choice of analytic
> methods.
Peter Nyikos
pnyikos
Apr 16, 2013, 10:54:47 PM4/16/13
to nyi...@bellsouth.net
On Apr 13, 8:43�pm, John Harshman <jharsh...@pacbell.net> wrote:
> On 4/12/13 6:58 PM, pnyikos wrote:
>
>
>
>
>
> > On Apr 10, 6:51 pm, John Harshman<jharsh...@pacbell.net> �wrote:
> >> On 4/10/13 3:30 PM, pnyikos wrote:
> >>> On Apr 10, 5:46 pm, John Harshman<jharsh...@pacbell.net> � �wrote:
> >>>> On 4/10/13 2:34 PM, alias Ernest Major wrote:
>
> >>>>> On 10/04/2013 21:56, Richard Norman wrote:
>
> >>>>>> I would prefer to return to the topic and so I ask John why
> >>>>>> "traditional systematists" are held in such low regard. Isn't this
> >>>>>> just a typical arrogance of the molecular people when there is
> >>>>>> disagreement about lineage? Is there a consensus about turtles or is
> >>>>>> there really some debate?
>
> >>>>> I think you'll find that it's not morphological vs molecular, but
> >>>>> traditional vs cladistic. (I'm all that sure of the low regard either.)
>
> >>> I think it is molecular vs. morphological that takes into account
> >>> extinct taxa, like *Eusthenopteron*.
>
> >> Then you are wrong. I think you're confusing the discussion of turtle
> >> cladistic relationships (not traditional systematics) with your ideas of
> >> close relationships to ancestors (traditional).
>
> > You misunderstood my intent, and maybe I misunderstood Ernest Major's
> > intent. �I was thinking of the last question Richard Norman asked, as
> > to where the controversy about turtle ancestry is centered.
> On 4/12/13 6:58 PM, pnyikos wrote:
>
>
>
>
>
> > On Apr 10, 6:51 pm, John Harshman<jharsh...@pacbell.net> �wrote:
> >> On 4/10/13 3:30 PM, pnyikos wrote:
> >>> On Apr 10, 5:46 pm, John Harshman<jharsh...@pacbell.net> � �wrote:
> >>>> On 4/10/13 2:34 PM, alias Ernest Major wrote:
>
> >>>>> On 10/04/2013 21:56, Richard Norman wrote:
>
> >>>>>> I would prefer to return to the topic and so I ask John why
> >>>>>> "traditional systematists" are held in such low regard. Isn't this
> >>>>>> just a typical arrogance of the molecular people when there is
> >>>>>> disagreement about lineage? Is there a consensus about turtles or is
> >>>>>> there really some debate?
>
> >>>>> I think you'll find that it's not morphological vs molecular, but
> >>>>> traditional vs cladistic. (I'm all that sure of the low regard either.)
>
> >>> I think it is molecular vs. morphological that takes into account
> >>> extinct taxa, like *Eusthenopteron*.
>
> >> Then you are wrong. I think you're confusing the discussion of turtle
> >> cladistic relationships (not traditional systematics) with your ideas of
> >> close relationships to ancestors (traditional).
>
> > You misunderstood my intent, and maybe I misunderstood Ernest Major's
> > intent. �I was thinking of the last question Richard Norman asked, as
>
> > A number of months ago I read an article by Gauthier in which he
> > described his latest morphological cladistic analysis, in which
> > extinct as well as extant species were incuded. �Among them was the
> > enigmatic *Eunotosaurus* as well as the earliest known Chelonian. �And
> > the result was that turtles were classed as anapsids.
>
> All true. But whatever does that have to do with traditional systematists?
Why ask me? Norman and Major were the ones who seemed to sense a
> > A number of months ago I read an article by Gauthier in which he
> > described his latest morphological cladistic analysis, in which
> > extinct as well as extant species were incuded. �Among them was the
> > enigmatic *Eunotosaurus* as well as the earliest known Chelonian. �And
> > the result was that turtles were classed as anapsids.
>
> All true. But whatever does that have to do with traditional systematists?
link between the two themes. See their words at the top.
> >>>>> The problem with traditional systematics is baseing classifications on a
> >>>>> small number of subjectively selected "key" characters, rather than on
> >>>>> all the evidence.
>
> >>> Largely true, but subjectivity is in the eye of the beholder.
>
> >>> For instance, I've said in the past that just the structure of the
> >>> shoulder girdle in the monotremes vs. marsupials and placentals should
> >>> outweigh a really hefty number of minor (especially molecular)
> >>> characters as far as what is the sister group of what.
>
> >>> The scapula of marsupials and placentals is essentially identical,
> >>> even to a raised median ridge, while that of monotremes is utterly
> >>> different.
>
> > And harking back to therapsids and even pelycosaurs. �There is a
> > series of fine drawings in Romer's _Vertebrate Paleontology showing
> > shoulder girdles of *Ophiacodon* (a pelycosaur), *Kannemeyeria* (a
> > dicynodont), the platypus, and the Virginia opossum. �The first three
> > resemble each other far more than the last two resemble each other,
> > even when the extra elements of the first three shoulder girdles are
> > disregarded.
>
> You're responding to yourself two layers back again.
home. The second time around, I had it to refer to.
I'd love to just post an url for diagrams like that, but I haven't
found one yet.
> Try to refrain.
Below, you interrupted me after a colon, forcing me to either
"respond to myself" or to refrain from replying to what you say next
until all I wrote earlier had transpired--or to do what I did, which
was to simply leave the reader guessing for a long time about what I
had been referring to.
> >>> � Also there is at least one extra element in the monotreme
> >>> shoulder girdle.
>
> > Two. The interclavicle and the procoracoid.
[counterproductive haggling snipped]
>
> > Two. The interclavicle and the procoracoid.
> >>>> That wasn't quite the "traditional" that Peter was talking about. He was
> >>>> referring, essentially, to gradism.
>
> >>> If you are referring to grades as in "subholostean grade" and
> >>> "holostean grade" you are dead wrong. �That is the only example of
> >>> polyphyletic taxon that Romer gave in _Vertebrate Paleontology_, the
> >>> example of both the bowfin and the gar being in Holostei.
>
> > No reply to this from you, so I am still confused by what you mean by
> > "gradism." �For now I will assume you really meant to refer to what �I
> > described next:
> Not at all. Most grades are paraphyletic. "Reptile" is a grade.
> "Thecodont" is a grade. "Mammal-like reptile" is a grade. Traditional
> "Amphibia" is a grade.
paraphyletic taxa to me, no more, no less.
In Romer's section "Taxonomy and classification" in Chapter 1 of
_Vertebrate Paleontology_ he never uses the word "grade," and the only
mention of grades I know of in the whole book is the one I've given
above, and he even calls the subholostean grade "an artificial
assemblage".
> And so on. Traditional systematists, including
> Simpson and Romer, liked grades.
type classifications. The latter is as close as he gets to grades,
and it does not seem close at all:
But when, for example, forms are discovered
seemingly ancestral to two distinct families,
or closely related to both their inclusion in
one or the other seems improper. Under such
circumstances the best solution seem to be
a "horizontal" cleavage, the erection of a stem
group, including the base from which the later
long-lived later families have been derived. (pp. 6-7)
Phylogeny remains uppermost in Romer's mind here, not some "grades" of
sophistication or whatever. That is exactly what I like about
paraphyletic taxa: they do justice to near-ancestry the way
cladistical classification cannot. By making reconstruction of the
purely hypothetical phylogeny from the clades the be-all and end-all
of classification, you lose out on the kind of understanding that well-
chosen paraphyletic taxa give you.
IMHO, Thecodontia was not well chosen, but the other three you name
were. In Romer's paralance, "old Reptilia" was a "stem group," with
ancestors in it for both Mammalia and Aves.
> >>> What I keep referring to is the traditional Linnean classification
> >>> whereby every organism had its own genus, family, order etc.
> >>> [including some intermediates like "suborder," "infraorder,"
> >>> "superfamily" in some cases]. �Cladistic classification is great for
> >>> extant taxa, but the further back in time one goes, the more
> >>> unsatisfactory it is.
Peter Nyikos
John Harshman
Apr 17, 2013, 1:02:11 AM4/17/13
to
replying to me.
>>>>> Also there is at least one extra element in the monotreme
>>>>> shoulder girdle.
>>
>>> Two. The interclavicle and the procoracoid.
>
> [counterproductive haggling snipped]
>
>>>>>> That wasn't quite the "traditional" that Peter was talking about. He was
>>>>>> referring, essentially, to gradism.
>>
>>>>> If you are referring to grades as in "subholostean grade" and
>>>>> "holostean grade" you are dead wrong. That is the only example of
>>>>> polyphyletic taxon that Romer gave in _Vertebrate Paleontology_, the
>>>>> example of both the bowfin and the gar being in Holostei.
>>
>>> No reply to this from you, so I am still confused by what you mean by
>>> "gradism." For now I will assume you really meant to refer to what I
>>> described next:
>
>> Not at all. Most grades are paraphyletic. "Reptile" is a grade.
>> "Thecodont" is a grade. "Mammal-like reptile" is a grade. Traditional
>> "Amphibia" is a grade.
>
> All this is new to me. All the examples you name are simply
> paraphyletic taxa to me, no more, no less.
> In Romer's section "Taxonomy and classification" in Chapter 1 of
> _Vertebrate Paleontology_ he never uses the word "grade," and the only
> mention of grades I know of in the whole book is the one I've given
> above, and he even calls the subholostean grade "an artificial
> assemblage".
meaning?
>> And so on. Traditional systematists, including
>> Simpson and Romer, liked grades.
>
> What Romer talks about in that section are "vertical" and "horizontal"
> type classifications. The latter is as close as he gets to grades,
> and it does not seem close at all:
>
> But when, for example, forms are discovered
> seemingly ancestral to two distinct families,
> or closely related to both their inclusion in
> one or the other seems improper. Under such
> circumstances the best solution seem to be
> a "horizontal" cleavage, the erection of a stem
> group, including the base from which the later
> long-lived later families have been derived. (pp. 6-7)
>
> Phylogeny remains uppermost in Romer's mind here, not some "grades" of
> sophistication or whatever. That is exactly what I like about
> paraphyletic taxa: they do justice to near-ancestry the way
> cladistical classification cannot.
or sophistication is intended. Grades are merely collections of species
that resemble each other in particular characters. When those characters
do not define clades, the grade is paraphyletic. As in "skin covered by
keratinous scales" defines a grade Reptilia.
> By making reconstruction of the
> purely hypothetical phylogeny from the clades the be-all and end-all
> of classification, you lose out on the kind of understanding that well-
> chosen paraphyletic taxa give you.
> IMHO, Thecodontia was not well chosen, but the other three you name
> were. In Romer's paralance, "old Reptilia" was a "stem group," with
> ancestors in it for both Mammalia and Aves.
John Harshman
Apr 17, 2013, 1:11:49 AM4/17/13
to
perchance searched GenBank for Sphenodon sequences?
> Right near the beginning of this thread, I said that the omission of
> Sphenodon was a serious flaw in the paper we started out discussing.
> A little later, the following exchange ensued between us:
>
> _________excerpt, you going first and last_______________
> >> There is no apparent correlation, based on all sorts of data in
> all
>>> sorts of taxa, between morphological and molecular evolutionary rates.
>
>> However, I am still curious to know what the situation is with
>> Sphenodon, as suggested in my first reply.
>
> You must remain curious until there's a Sphenodon genome or a
> substantial portion thereof.
> ============ end of excerpt===========
>
>> So "fragmentary information" is plenty. Just not that
>> fragmentary.
>
> And indeed, it overlaps with "substantial portion" in your
> vocabulary. I take it that not even a measly 10^4 base pairs of
> Sphenodon (the tuatara) have been sequenced in the 17 years since
> those blood proteins were sequenced.
hard to read. Please stop that sort of thing. A simple search of GenBank
finds over 2000 Sphenodon sequence records, of varying lengths, but
certainly many thousands of bases in toto. Perhaps you should check that.
Anyway, you would likely need a bigger sample to estimate a genome-wide
evolutionary rate than you would to estimate a phylogeny.
> In New Zealand between then and now (2001, if memory serves) I saw
> some live tuataras in captivity. Surely the sample needed to get lots
> of base pairs would have been small and harmless.
>> And NJ was probably an especially poor choice of analytic
>> methods.
>
> Still a legitimate part of cladistics, though, no?
asked for my view, I would say that it isn't.
pnyikos
Apr 17, 2013, 11:27:57 PM4/17/13
to nyi...@bellsouth.net
On Apr 17, 1:11�am, John Harshman <jharsh...@pacbell.net> wrote:
> On 4/16/13 6:49 PM, pnyikos wrote:
>
> > On Apr 16, 3:56 pm, John Harshman<jharsh...@pacbell.net> �wrote:
> >> On 4/16/13 12:44 PM, pnyikos wrote:
>
> >>> On Apr 15, 10:07 pm, John Harshman<jharsh...@pacbell.net> � �wrote:
> >>>> On 4/15/13 9:26 AM, pnyikos wrote:
> >>>>> The moral is that, �as long as we have only fragmentary information
> >>>>> about the Sphenodon genome, its relationships are up in the air.
>
> > [snip]
>
> >>>> No, the moral is that a neighbor-joining analysis of a short
> >>>> DNA sequence isn't all that reliable.
>
> >>> Do you have any reason whatsoever for thinking *my* moral is false?
> >>> Please keep my conditional "as long as" in mind.
>
> >> Yes. The distance between a small, single gene and an entire genome is
> >> big enough that the lower reaches of that distance would still be quite
> >> sufficient for practical certainty about relationships.
I'm not sure Gauthier and Laurin would agree with the figures you give
> On 4/16/13 6:49 PM, pnyikos wrote:
>
> > On Apr 16, 3:56 pm, John Harshman<jharsh...@pacbell.net> �wrote:
> >> On 4/16/13 12:44 PM, pnyikos wrote:
>
> >>> On Apr 15, 10:07 pm, John Harshman<jharsh...@pacbell.net> � �wrote:
> >>>> On 4/15/13 9:26 AM, pnyikos wrote:
> >>>>> The moral is that, �as long as we have only fragmentary information
> >>>>> about the Sphenodon genome, its relationships are up in the air.
>
> > [snip]
>
> >>>> No, the moral is that a neighbor-joining analysis of a short
> >>>> DNA sequence isn't all that reliable.
>
> >>> Do you have any reason whatsoever for thinking *my* moral is false?
> >>> Please keep my conditional "as long as" in mind.
>
> >> Yes. The distance between a small, single gene and an entire genome is
> >> big enough that the lower reaches of that distance would still be quite
> >> sufficient for practical certainty about relationships.
below. Read on
> >> That is, the
> >> paper you reference had perhaps a few times 10^2 bases, the genome is a
> >> few times 10^9, and a few times 10^3 or 10^4 would almost certainly do
> >> the trick.
the following question:
> > These numbers are all fine and dandy, if correct [where can I read up
> > on these things?]
> > in mind the current state of fragmentariness of what we know of the
> > Sphenodon genome.
>
> What do you know of the current state of "fragmentariness"? Have you
> perchance searched GenBank for Sphenodon sequences?
No. What am I supposed to do with them once I find them, call them to
> > Sphenodon genome.
>
> What do you know of the current state of "fragmentariness"? Have you
> perchance searched GenBank for Sphenodon sequences?
the attention of Gauthier and Laurin? Here is what they write about
the state of the art back in 1999, and seem to be oblivious to how
much progress (if any) has been made since then:
"Many gene sequences of birds and mammals exist, but the relatively
small number of sequences from representatives of other amniote
lineages, especially tuataras (Sphenodon) and turtles, has hindered
the estimation of a robust molecular phylogeny for all major groups of
living amniotes. This is reflected by the low resolution of the
molecular phylogeny obtained by Hedges & Poling (1999) when Sphenodon
(using only sequences of genes available in Sphenodon) was included."
--http://tolweb.org/Amniota/14990
There follows a little tree showing Sphenodon, Testudines, Crocodylia
and Aves in a tetrachotomy, with Squamates as the sister group.
>
>
>
> > Right near the beginning of this thread, I said that the omission of
> > Sphenodon was a serious flaw in the paper we started out discussing.
> > A little later, the following exchange ensued between us:
>
> > _________excerpt, you going first and last_______________
> > � >> �There is no apparent correlation, based on all sorts of data in
> > all
> >>> sorts of taxa, between morphological and molecular evolutionary rates.
>
> >> However, I am still curious to know what the situation is with
> >> Sphenodon, as suggested in my first reply.
>
> > You must remain curious until there's a Sphenodon genome or a
> > substantial portion thereof.
> > ============ end of excerpt===========
>
> >> So "fragmentary information" is plenty. Just not that
> >> fragmentary.
>
> > And indeed, it overlaps with "substantial portion" in your
> > vocabulary. �I take it that not even a measly 10^4 base pairs of
> > Sphenodon (the tuatara) have been sequenced in the 17 years since
> > those blood proteins were sequenced.
>
> Your habit of random flashbacks and such is once again making your posts
> hard to read. Please stop that sort of thing.
time with things you write at another time. If I were as
knowledgeable about systematics as Gauthier, or even Cal King, I might
have an easier time following you.
> A simple search of GenBank
> finds over 2000 Sphenodon sequence records, of varying lengths, but
> certainly many thousands of bases in toto. Perhaps you should check that.
>
> Anyway, you would likely need a bigger sample to estimate a genome-wide
> evolutionary rate than you would to estimate a phylogeny.
comment. My bad.
> > In New Zealand between then and now (2001, if memory serves) I saw
> > some live tuataras in captivity. �Surely the sample needed to get lots
> > of base pairs would have been small and harmless.
>
> That would certainly not be the problem, if any.
>
> >> And NJ was probably an especially poor choice of analytic
> >> methods.
>
> > Still a legitimate part of cladistics, though, no?
>
> Depends on what you mean by "legitimate" and "cladistics".
by cladophiles like yourself, the way paraphyletic taxa are banished.
>�But if you
> asked for my view, I would say that it isn't.
Don't systematists use NJ any more?
Peter Nyikos
pnyikos
Apr 17, 2013, 11:57:48 PM4/17/13
to nyi...@bellsouth.net
On Apr 17, 1:02�am, John Harshman <jharsh...@pacbell.net> wrote:
> On 4/16/13 7:54 PM, pnyikos wrote:
> > On Apr 13, 8:43 pm, John Harshman<jharsh...@pacbell.net> �wrote:
> >> On 4/12/13 6:58 PM, pnyikos wrote:
>
> >>> On Apr 10, 6:51 pm, John Harshman<jharsh...@pacbell.net> � �wrote:
> On 4/16/13 7:54 PM, pnyikos wrote:
> > On Apr 13, 8:43 pm, John Harshman<jharsh...@pacbell.net> �wrote:
> >> On 4/12/13 6:58 PM, pnyikos wrote:
>
> >>> On Apr 10, 6:51 pm, John Harshman<jharsh...@pacbell.net> � �wrote:
> >> You're responding to yourself two layers back again.
>
> > The first time around, I was at my office, and Romer's book is at
> > home. �The second time around, I had it to refer to.
>
> > I'd love to just post an url for diagrams like that, but I haven't
> > found one yet.
>
> >> Try to refrain.
>
> > Below, you interrupted me after a colon, �forcing me to either
> > "respond to myself" or to refrain from replying to what you say next
> > until all I wrote earlier had transpired--or to do what I did, which
> > was to simply leave the reader guessing for a long time about what I
> > had been referring to.
>
> Or you could reply to me right after what I say, in the post that
> replying to me.
Except for that enigmatic last clause, you are paraphrasing the third
>
> > The first time around, I was at my office, and Romer's book is at
> > home. �The second time around, I had it to refer to.
>
> > I'd love to just post an url for diagrams like that, but I haven't
> > found one yet.
>
> >> Try to refrain.
>
> > Below, you interrupted me after a colon, �forcing me to either
> > "respond to myself" or to refrain from replying to what you say next
> > until all I wrote earlier had transpired--or to do what I did, which
> > was to simply leave the reader guessing for a long time about what I
> > had been referring to.
>
> Or you could reply to me right after what I say, in the post that
> replying to me.
option I gave; and indeed, that is the option I followed.
> >>>>> � �Also there is at least one extra element in the monotreme
> >>>>> shoulder girdle.
>
> >>> Two. The interclavicle and the procoracoid.
>
> > [counterproductive haggling snipped]
>
> >>>>>> That wasn't quite the "traditional" that Peter was talking about. He was
> >>>>>> referring, essentially, to gradism.
>
> >>>>> If you are referring to grades as in "subholostean grade" and
> >>>>> "holostean grade" you are dead wrong. �That is the only example of
> >>>>> polyphyletic taxon that Romer gave in _Vertebrate Paleontology_, the
> >>>>> example of both the bowfin and the gar being in Holostei.
>
> >>> No reply to this from you, so I am still confused by what you mean by
> >>> "gradism." �For now I will assume you really meant to refer to what �I
> >>> described next:
>
> >> Not at all. Most grades are paraphyletic. "Reptile" is a grade.
> >> "Thecodont" is a grade. "Mammal-like reptile" is a grade. Traditional
> >> "Amphibia" is a grade.
>
> > All this is new to me. �All the examples you name are simply
> > paraphyletic taxa to me, no more, no less.
>
> paraphyletic taxon = grade, mostly.
Sorry, I don't buy it.
>
> >>> Two. The interclavicle and the procoracoid.
>
> > [counterproductive haggling snipped]
>
> >>>>>> That wasn't quite the "traditional" that Peter was talking about. He was
> >>>>>> referring, essentially, to gradism.
>
> >>>>> If you are referring to grades as in "subholostean grade" and
> >>>>> "holostean grade" you are dead wrong. �That is the only example of
> >>>>> polyphyletic taxon that Romer gave in _Vertebrate Paleontology_, the
> >>>>> example of both the bowfin and the gar being in Holostei.
>
> >>> No reply to this from you, so I am still confused by what you mean by
> >>> "gradism." �For now I will assume you really meant to refer to what �I
> >>> described next:
>
> >> Not at all. Most grades are paraphyletic. "Reptile" is a grade.
> >> "Thecodont" is a grade. "Mammal-like reptile" is a grade. Traditional
> >> "Amphibia" is a grade.
>
> > All this is new to me. �All the examples you name are simply
> > paraphyletic taxa to me, no more, no less.
>
> paraphyletic taxon = grade, mostly.
> > In Romer's section "Taxonomy and classification" in Chapter 1 of
> > _Vertebrate Paleontology_ he never uses the word "grade," and the only
> > mention of grades I know of in the whole book is the one I've given
> > above, and he even calls the subholostean grade "an artificial
> > assemblage".
>
> So what? Just because he doesn't use the word, we can't recognize the
> meaning?
Chondrostei, Holostei, and Teleostei:
"These terms should be regarded as progressive
grades of development within the Actinopterygii
rather than true phyletic units; for, as will be seen,
the holostean level of organization was surely
attained by more than one group of primitive forms,
and the teleosts may be similarly polyphyletic."
_Vertebrate Paleontology_, 2nd ed., p. 87
Note, "polyphyletic," not "paraphyletic." The latter he deemed to be
"true phyletic units" based on my extensive perusal of his book. He
goes on to explain why he bothers with these grades at all:
"However, this type of classification, even if not
entirely naural, is one which it is best to preserve
until our knowledge of the details of the complex
evolutionary history of the ray-finned fishes is much
more adequate than is the case at present."
> >> And so on. Traditional systematists, including
> >> Simpson and Romer, liked grades.
that you've been relying on hearsay rather than studying what they
actually wrote, at least in Romer's case.
> > What Romer talks about in that section are "vertical" and "horizontal"
> > type classifications. �The latter is as close as he gets to grades,
> > and it does not seem close at all:
>
> > � � But when, for example, forms are discovered
> > � � seemingly ancestral to two distinct families,
> > � � or closely related to both their inclusion in
> > � � one or the other seems improper. Under such
> > � � circumstances the best solution seem to be
> > � � a "horizontal" cleavage, the erection of a stem
> > � � group, including the base from which the later
> > � � long-lived later families have been derived. (pp. 6-7)
>
> > Phylogeny remains uppermost in Romer's mind here, not some "grades" of
> > sophistication or whatever. That is exactly what I like about
> > paraphyletic taxa: they do justice to near-ancestry the way
> > cladistical classification cannot.
>
> I believe you misunderstand the meaning of "grade". No measure of value
> or sophistication is intended. Grades are merely collections of species
> that resemble each other in particular characters. When those characters
> do not define clades, the grade is paraphyletic.
old traditional paleontology texts to be littered with polyphyletic
"taxa" if things were as you describe. They aren't.
> As in "skin covered by
> keratinous scales" defines a grade Reptilia.
keratin. Then Romer would have relied on <gasp> phylogeny.
> > By making reconstruction of the
> > purely hypothetical phylogeny from the clades the be-all and end-all
> > of classification, you lose out on the kind of understanding that well-
> > chosen paraphyletic taxa give you.
>
> That isn't understanding.
> > IMHO, Thecodontia was not well chosen, but the other three you name
> > were. �In Romer's paralance, "old Reptilia" was a "stem group," with
> > ancestors in it for both Mammalia and Aves.
>
> Yep. A grade.
alias Ernest Major
Apr 18, 2013, 4:58:31 AM4/18/13
to
relationships. Neighbour joining uses a measure of similarity to infer a
tree, so on the face of it it is not a cladistic method. It also seems
likely to be vulnerable to artifacts resulting from differences in
evolutionary rates between lineages.
>
>> But if you
>> asked for my view, I would say that it isn't.
>
> Don't systematists use NJ any more?
Google Scholar is your friend.
>> But if you
>> asked for my view, I would say that it isn't.
>
> Don't systematists use NJ any more?
Neighbour-joining runs in polynomial time (typically O(n^3), but there's
a recentish variant which is O(n^2). Maximum-parsimony is, fide
Wikipedia, NP-hard. So on the face of it there would be a range of
problems for which neighbour-joining is feasible, and maximum-parsimony
isn't.
>
> Peter Nyikos
>
--
alias Ernest Major
John Harshman
Apr 18, 2013, 9:51:27 AM4/18/13
to
On 4/17/13 8:57 PM, pnyikos wrote:
> On Apr 17, 1:02 am, John Harshman<jharsh...@pacbell.net> wrote:
>> On 4/16/13 7:54 PM, pnyikos wrote:
>>> On Apr 13, 8:43 pm, John Harshman<jharsh...@pacbell.net> wrote:
>>>> On 4/12/13 6:58 PM, pnyikos wrote:
>>
>>>>> On Apr 10, 6:51 pm, John Harshman<jharsh...@pacbell.net> wrote:
>
>>>> You're responding to yourself two layers back again.
>>
>>> The first time around, I was at my office, and Romer's book is at
>>> home. The second time around, I had it to refer to.
>>
>>> I'd love to just post an url for diagrams like that, but I haven't
>>> found one yet.
>>
>>>> Try to refrain.
>>
>>> Below, you interrupted me after a colon, forcing me to either
>>> "respond to myself" or to refrain from replying to what you say next
>>> until all I wrote earlier had transpired--or to do what I did, which
>>> was to simply leave the reader guessing for a long time about what I
>>> had been referring to.
>>
>> Or you could reply to me right after what I say, in the post that
>> replying to me.
>
> Except for that enigmatic last clause, you are paraphrasing the third
> option I gave; and indeed, that is the option I followed.
No it isn't. You replied to posts several layers back and attached old
> On Apr 17, 1:02 am, John Harshman<jharsh...@pacbell.net> wrote:
>> On 4/16/13 7:54 PM, pnyikos wrote:
>>> On Apr 13, 8:43 pm, John Harshman<jharsh...@pacbell.net> wrote:
>>>> On 4/12/13 6:58 PM, pnyikos wrote:
>>
>>>>> On Apr 10, 6:51 pm, John Harshman<jharsh...@pacbell.net> wrote:
>
>>>> You're responding to yourself two layers back again.
>>
>>> The first time around, I was at my office, and Romer's book is at
>>> home. The second time around, I had it to refer to.
>>
>>> I'd love to just post an url for diagrams like that, but I haven't
>>> found one yet.
>>
>>>> Try to refrain.
>>
>>> Below, you interrupted me after a colon, forcing me to either
>>> "respond to myself" or to refrain from replying to what you say next
>>> until all I wrote earlier had transpired--or to do what I did, which
>>> was to simply leave the reader guessing for a long time about what I
>>> had been referring to.
>>
>> Or you could reply to me right after what I say, in the post that
>> replying to me.
>
> Except for that enigmatic last clause, you are paraphrasing the third
> option I gave; and indeed, that is the option I followed.
material to it.
>>>>>>> Also there is at least one extra element in the monotreme
>>>>>>> shoulder girdle.
>>
>>>>> Two. The interclavicle and the procoracoid.
>>
>>> [counterproductive haggling snipped]
>>
>>>>>>>> That wasn't quite the "traditional" that Peter was talking about. He was
>>>>>>>> referring, essentially, to gradism.
>>
>>>>>>> If you are referring to grades as in "subholostean grade" and
>>>>>>> "holostean grade" you are dead wrong. That is the only example of
>>>>>>> polyphyletic taxon that Romer gave in _Vertebrate Paleontology_, the
>>>>>>> example of both the bowfin and the gar being in Holostei.
>>
>>>>> No reply to this from you, so I am still confused by what you mean by
>>>>> "gradism." For now I will assume you really meant to refer to what I
>>>>> described next:
>>
>>>> Not at all. Most grades are paraphyletic. "Reptile" is a grade.
>>>> "Thecodont" is a grade. "Mammal-like reptile" is a grade. Traditional
>>>> "Amphibia" is a grade.
>>
>>> All this is new to me. All the examples you name are simply
>>> paraphyletic taxa to me, no more, no less.
>>
>> paraphyletic taxon = grade, mostly.
>
> Sorry, I don't buy it.
>>> In Romer's section "Taxonomy and classification" in Chapter 1 of
>>> _Vertebrate Paleontology_ he never uses the word "grade," and the only
>>> mention of grades I know of in the whole book is the one I've given
>>> above, and he even calls the subholostean grade "an artificial
>>> assemblage".
>>
>> So what? Just because he doesn't use the word, we can't recognize the
>> meaning?
>
> From what? Not from passages like the following paragraph about
> Chondrostei, Holostei, and Teleostei:
>
> "These terms should be regarded as progressive
> grades of development within the Actinopterygii
> rather than true phyletic units; for, as will be seen,
> the holostean level of organization was surely
> attained by more than one group of primitive forms,
> and the teleosts may be similarly polyphyletic."
> _Vertebrate Paleontology_, 2nd ed., p. 87
>
> Note, "polyphyletic," not "paraphyletic." The latter he deemed to be
> "true phyletic units" based on my extensive perusal of his book. He
> goes on to explain why he bothers with these grades at all:
>
> "However, this type of classification, even if not
> entirely naural, is one which it is best to preserve
> until our knowledge of the details of the complex
> evolutionary history of the ray-finned fishes is much
> more adequate than is the case at present."
fact may have been coined after he wrote. Nor is his understanding of
the word "grade" -- or anything else -- necessarily the one we should
take as our standard. I'm not arguing about Romer's definition. I'm
arguing about whether the groups he liked fit our modern definition of
"grade".
>>>> And so on. Traditional systematists, including
>>>> Simpson and Romer, liked grades.
>
> Judging from what I've quoted from him above and below, it would seem
> that you've been relying on hearsay rather than studying what they
> actually wrote, at least in Romer's case.
relying on our modern understanding of what "grade" means, not whether
Romer liked things he himself called "grades".
group would be based on the presence of certain character states, and
would exclude those with more derived states. That's what a grade is.
>>> By making reconstruction of the
>>> purely hypothetical phylogeny from the clades the be-all and end-all
>>> of classification, you lose out on the kind of understanding that well-
>>> chosen paraphyletic taxa give you.
>>
>> That isn't understanding.
>
> The word "That" hangs in mid-air.
taxa give you".
>
>>> IMHO, Thecodontia was not well chosen, but the other three you name
>>> were. In Romer's paralance, "old Reptilia" was a "stem group," with
>>> ancestors in it for both Mammalia and Aves.
>>
>> Yep. A grade.
>
> Sorry, I don't buy it.
We appear to be arguing over whose definition of "grade" to use. You
>>> were. In Romer's paralance, "old Reptilia" was a "stem group," with
>>> ancestors in it for both Mammalia and Aves.
>>
>> Yep. A grade.
>
> Sorry, I don't buy it.
like Romer's implied definition, and I use the one that is pretty much
universal in present usage.
John Harshman
Apr 18, 2013, 10:01:56 AM4/18/13
to
On 4/17/13 8:27 PM, pnyikos wrote:
> On Apr 17, 1:11 am, John Harshman<jharsh...@pacbell.net> wrote:
>> On 4/16/13 6:49 PM, pnyikos wrote:
>>
>>> On Apr 16, 3:56 pm, John Harshman<jharsh...@pacbell.net> wrote:
>>>> On 4/16/13 12:44 PM, pnyikos wrote:
>>
>>>>> On Apr 15, 10:07 pm, John Harshman<jharsh...@pacbell.net> wrote:
>>>>>> On 4/15/13 9:26 AM, pnyikos wrote:
>
>>>>>>> The moral is that, as long as we have only fragmentary information
>>>>>>> about the Sphenodon genome, its relationships are up in the air.
>>
>>> [snip]
>>
>>>>>> No, the moral is that a neighbor-joining analysis of a short
>>>>>> DNA sequence isn't all that reliable.
>>
>>>>> Do you have any reason whatsoever for thinking *my* moral is false?
>>>>> Please keep my conditional "as long as" in mind.
>>
>>>> Yes. The distance between a small, single gene and an entire genome is
>>>> big enough that the lower reaches of that distance would still be quite
>>>> sufficient for practical certainty about relationships.
>
> I'm not sure Gauthier and Laurin would agree with the figures you give
> below. Read on
Arrgh. Please stop replying to statements several layers down.
> On Apr 17, 1:11 am, John Harshman<jharsh...@pacbell.net> wrote:
>> On 4/16/13 6:49 PM, pnyikos wrote:
>>
>>> On Apr 16, 3:56 pm, John Harshman<jharsh...@pacbell.net> wrote:
>>>> On 4/16/13 12:44 PM, pnyikos wrote:
>>
>>>>> On Apr 15, 10:07 pm, John Harshman<jharsh...@pacbell.net> wrote:
>>>>>> On 4/15/13 9:26 AM, pnyikos wrote:
>
>>>>>>> The moral is that, as long as we have only fragmentary information
>>>>>>> about the Sphenodon genome, its relationships are up in the air.
>>
>>> [snip]
>>
>>>>>> No, the moral is that a neighbor-joining analysis of a short
>>>>>> DNA sequence isn't all that reliable.
>>
>>>>> Do you have any reason whatsoever for thinking *my* moral is false?
>>>>> Please keep my conditional "as long as" in mind.
>>
>>>> Yes. The distance between a small, single gene and an entire genome is
>>>> big enough that the lower reaches of that distance would still be quite
>>>> sufficient for practical certainty about relationships.
>
> I'm not sure Gauthier and Laurin would agree with the figures you give
> below. Read on
>>>> That is, the
>>>> paper you reference had perhaps a few times 10^2 bases, the genome is a
>>>> few times 10^9, and a few times 10^3 or 10^4 would almost certainly do
>>>> the trick.
>
> Even though it was in brackets, I would have appreciated an answer to
> the following question:
>>> These numbers are all fine and dandy, if correct [where can I read up
>>> on these things?]
resolve a node? I can't direct you to a single source.
>>> But I did not quantify "only fragmentary" and I had
>>> in mind the current state of fragmentariness of what we know of the
>>> Sphenodon genome.
>>
>> What do you know of the current state of "fragmentariness"? Have you
>> perchance searched GenBank for Sphenodon sequences?
>
> No. What am I supposed to do with them once I find them, call them to
> the attention of Gauthier and Laurin?
citation to the publication in which it's used. And thus you could find
papers in which Sphenodon sequence is used in investigating amniote
phylogeny.
> Here is what they write about
> the state of the art back in 1999, and seem to be oblivious to how
> much progress (if any) has been made since then:
>
> "Many gene sequences of birds and mammals exist, but the relatively
> small number of sequences from representatives of other amniote
> lineages, especially tuataras (Sphenodon) and turtles, has hindered
> the estimation of a robust molecular phylogeny for all major groups of
> living amniotes. This is reflected by the low resolution of the
> (using only sequences of genes available in Sphenodon) was included."
> --http://tolweb.org/Amniota/14990
>
> There follows a little tree showing Sphenodon, Testudines, Crocodylia
> and Aves in a tetrachotomy, with Squamates as the sister group.
That was indeed a long time ago.
> --http://tolweb.org/Amniota/14990
>
> There follows a little tree showing Sphenodon, Testudines, Crocodylia
> and Aves in a tetrachotomy, with Squamates as the sister group.
>>> Right near the beginning of this thread, I said that the omission of
>>> Sphenodon was a serious flaw in the paper we started out discussing.
>>> A little later, the following exchange ensued between us:
>>
>>> _________excerpt, you going first and last_______________
>>> >> There is no apparent correlation, based on all sorts of data in
>>> all
>>>>> sorts of taxa, between morphological and molecular evolutionary rates.
>>
>>>> However, I am still curious to know what the situation is with
>>>> Sphenodon, as suggested in my first reply.
>>
>>> You must remain curious until there's a Sphenodon genome or a
>>> substantial portion thereof.
>>> ============ end of excerpt===========
>>
>>>> So "fragmentary information" is plenty. Just not that
>>>> fragmentary.
>>
>>> And indeed, it overlaps with "substantial portion" in your
>>> vocabulary. I take it that not even a measly 10^4 base pairs of
>>> Sphenodon (the tuatara) have been sequenced in the 17 years since
>>> those blood proteins were sequenced.
>>
>> Your habit of random flashbacks and such is once again making your posts
>> hard to read. Please stop that sort of thing.
>
> Sorry, it is sometimes necessary to compare things you write at one
> time with things you write at another time. If I were as
> knowledgeable about systematics as Gauthier, or even Cal King, I might
> have an easier time following you.
this case no extensive quote was needed. All you had to do was ask about
what you thought was a discrepancy between my claim that you would need
only fragments for one purpose but a substantial portion of the genome
for another purpose.
>> A simple search of GenBank
>> finds over 2000 Sphenodon sequence records, of varying lengths, but
>> certainly many thousands of bases in toto. Perhaps you should check that.
>>
>> Anyway, you would likely need a bigger sample to estimate a genome-wide
>> evolutionary rate than you would to estimate a phylogeny.
>
> True, I wasn't paying enough attention to the context of your
> comment. My bad.
>
>>> In New Zealand between then and now (2001, if memory serves) I saw
>>> some live tuataras in captivity. Surely the sample needed to get lots
>>> of base pairs would have been small and harmless.
>>
>> That would certainly not be the problem, if any.
>>
>>>> And NJ was probably an especially poor choice of analytic
>>>> methods.
>>
>>> Still a legitimate part of cladistics, though, no?
>>
>> Depends on what you mean by "legitimate" and "cladistics".
>
> Anything used in systematics that isn't banished to the outer darkness
> by cladophiles like yourself, the way paraphyletic taxa are banished.
though cladophiles sufficiently unlike myself might.
>> But if you
>> asked for my view, I would say that it isn't.
>
> Don't systematists use NJ any more?
a best-fit, least squares tree. But it has problems. It's used mostly by
molecular biologists and others who don't want to do a real phylogenetic
analysis.
pnyikos
Apr 18, 2013, 9:37:50 PM4/18/13
to nyi...@bellsouth.net
On Apr 18, 9:51 am, John Harshman <jharsh...@pacbell.net> wrote:
> On 4/17/13 8:57 PM, pnyikos wrote:
> > On Apr 17, 1:02 am, John Harshman<jharsh...@pacbell.net> wrote:
> >> On 4/16/13 7:54 PM, pnyikos wrote:
> >>> On Apr 13, 8:43 pm, John Harshman<jharsh...@pacbell.net> wrote:
> >>>> On 4/12/13 6:58 PM, pnyikos wrote:
>
> >>>>> On Apr 10, 6:51 pm, John Harshman<jharsh...@pacbell.net> wrote:
preparing a talk I will be giving at a seminar at UNC-Charlotte on
Saturday, so this may be my last post to this thread till Monday.
But I don't want that to dissuade any reader from replying to my posts
whenever it is best for them.
> >>>>>>>> That wasn't quite the "traditional" that Peter was talking about. He was
> >>>>>>>> referring, essentially, to gradism.
>
> >>>>>>> If you are referring to grades as in "subholostean grade" and
> >>>>>>> "holostean grade" you are dead wrong. That is the only example of
> >>>>>>> polyphyletic taxon that Romer gave in _Vertebrate Paleontology_, the
> >>>>>>> example of both the bowfin and the gar being in Holostei.
>
> >>>>> No reply to this from you, so I am still confused by what you mean by
> >>>>> "gradism." For now I will assume you really meant to refer to what I
> >>>>> described next:
>
> >>>> Not at all. Most grades are paraphyletic. "Reptile" is a grade.
> >>>> "Thecodont" is a grade. "Mammal-like reptile" is a grade. Traditional
> >>>> "Amphibia" is a grade.
>
> >>> All this is new to me. All the examples you name are simply
> >>> paraphyletic taxa to me, no more, no less.
>
> >> paraphyletic taxon = grade, mostly.
>
> > Sorry, I don't buy it.
>
> I can't help that.
should mean to us, and where you get the idea that they were
extensively used by 20th century paleontologists before cladistics
came along.
an illustration of a phylogenetic tree that is subjected to a
"horizontal cleavage" [see what I quoted from him below] and the part
below the line in a horizontal cleavage is a paraphyletic taxon by
definition.
> which in
> fact may have been coined after he wrote. Nor is his understanding of
> the word "grade" -- or anything else -- necessarily the one we should
> take as our standard. I'm not arguing about Romer's definition. I'm
> arguing about whether the groups he liked fit our modern definition of
> "grade".
under the delusion that all paraphyletic taxa are produced from a
clade by removing ONE clade when he had the example of Reptilia, with
both Aves and Mammalia removed, to tell him otherwise?
I do believe, though, that he was more on the right track than whoever
it was that came up with the formula that "monophyletic taxa are
defined by apomorphies, paraphyletic by plesimorphies, and
polyphyletic by similarity." Romer et. al. produced their
"paraphyletic" taxa via "horizontal cleavage" or else they inherited
them from earlier systematists, threw out the ones that were
polyphyletic (with a few exceptions like Holostei, which were looked
upon as "necessary evils") and hung on to the "true phyletic groups,"
both paraphyletic and holophyletic.
> >>>> And so on. Traditional systematists, including
> >>>> Simpson and Romer, liked grades.
>
> > Judging from what I've quoted from him above and below, it would seem
> > that you've been relying on hearsay rather than studying what they
> > actually wrote, at least in Romer's case.
>
> There you go again, responding to old statements. Please stop.
refuses to allow testimony that would prove innocence or guilt because
the lawyer who wants to present it learned about new evidence after he
already gave one speech on the subject.
And I stand by what I wrote and you ignored.
> And I'm
> relying on our modern understanding of what "grade" means, not whether
> Romer liked things he himself called "grades".
modern folks call grades. I see NO evidence that he liked them, and
the quotes I gave from him strongly suggest that he disliked them.
> and
> would exclude those with more derived states. That's what a grade is.
See above.
Remainder deleted, to be replied to later (probably next week).
Peter Nyikos
Richard Norman
Apr 18, 2013, 10:51:35 PM4/18/13
to
but I can tell you absolutely that the modern concept of "clade" is
what was introduced by Hennig in the mid 1960's. The introduction of
purely (or relatively so) cladistic thinking within biology did not
become generally accepted until the '70s or even later. Many of the
groupings by pre-cladist biologicsts were based on possessing some
shared characteristic, especially "level of organization" which is, in
fact, exactly how a "grade" is defined. Wikipedia: In alpha taxonomy,
a grade refers to a taxon united by a level of morphological or
physiological complexity. The term was coined by British biologist
Julian Huxley, to contrast with clade, a strictly phylogenetic unit.
http://en.wikipedia.org/wiki/Evolutionary_grade
Note: Huxley used the concept of clade somehwat less than a decade
before Hennig but it didn't really catch on until people started
arguing about Hennig's ideas.
That source goes on to say: "Thus, traditional palaeontological works
are often using evolutionary grades as formal or informal taxa,
including examples such as Labyrinthodonts, Anapsids, Synapsids,
Dinosaurs, Ammonites, Eurypterid, Lobopodes and many of the more well
known taxa of human evolution. Organizing organisms into grades rather
than strict clades can also be very useful to understand the
evolutionary sequence behind major diversification of both animals[6]
and plants."
Interestingly, the Wikipedia page on cladistics uses, as an example,
the disputed placement of turtles within and outsied the diapsids.
http://en.wikipedia.org/wiki/Cladistics
Further, I really don't understand why you argue so intensely about
what Romer might have written in an era when molecular genetic data
about phylogenetic relationships was essentially completely absent.
John Harshman
Apr 19, 2013, 10:44:56 AM4/19/13
to
"grade" and substitute "paraphyletic group", which means the same thing.
Nor do I want to argue about the meaning of "grade", since it's
peripheral to that point. (Though that's exactly what "grade" does mean.)
>> which in
>> fact may have been coined after he wrote. Nor is his understanding of
>> the word "grade" -- or anything else -- necessarily the one we should
>> take as our standard. I'm not arguing about Romer's definition. I'm
>> arguing about whether the groups he liked fit our modern definition of
>> "grade".
>
> And who is the authority on what that definition is? Ferris, who was
> under the delusion that all paraphyletic taxa are produced from a
> clade by removing ONE clade when he had the example of Reptilia, with
> both Aves and Mammalia removed, to tell him otherwise?
> I do believe, though, that he was more on the right track than whoever
> it was that came up with the formula that "monophyletic taxa are
> defined by apomorphies, paraphyletic by plesimorphies, and
> polyphyletic by similarity."
past triumphs despite their marginal connections to the case at hand. We
aren't arguing about the definition of paraphyly. And I'm not sure who
"Ferris" is, unless you refer to Steve Farris. Anyway, we all here agree
on what paraphyly is.
> Romer et. al. produced their
> "paraphyletic" taxa via "horizontal cleavage" or else they inherited
> them from earlier systematists, threw out the ones that were
> polyphyletic (with a few exceptions like Holostei, which were looked
> upon as "necessary evils") and hung on to the "true phyletic groups,"
> both paraphyletic and holophyletic.
>>>>>> And so on. Traditional systematists, including
>>>>>> Simpson and Romer, liked grades.
>>
>>> Judging from what I've quoted from him above and below, it would seem
>>> that you've been relying on hearsay rather than studying what they
>>> actually wrote, at least in Romer's case.
>>
>> There you go again, responding to old statements. Please stop.
>
> Cut it out, Harshman. You are sounding like a capricious judge who
> refuses to allow testimony that would prove innocence or guilt because
> the lawyer who wants to present it learned about new evidence after he
> already gave one speech on the subject.
>
> And I stand by what I wrote and you ignored.
have described are grades or, if that word is contentious, paraphyletic
taxa. And that was my point.
>> And I'm
>> relying on our modern understanding of what "grade" means, not whether
>> Romer liked things he himself called "grades".
>
> Irrelevant to the real issue, which is whether Romer liked what you
> modern folks call grades. I see NO evidence that he liked them, and
> the quotes I gave from him strongly suggest that he disliked them.
accept it on occasion). You have shown that he liked paraphyly just fine.
more. While phylogeny is estimated using characters, a clade is not
based on particular character states, and in fact there may be no state
universal within a clade.
>> and
>> would exclude those with more derived states. That's what a grade is.
>
> Romer's practice of systematics does not fit that Procrustean bed.
> See above.
taxa with derived states.
pnyikos
Apr 22, 2013, 1:30:32 PM4/22/13
to nyi...@bellsouth.net
Google groups still supports sci.bio.systematics, so I've added it
just in case someone interested in systematics chances across this
thread before it becomes buried under spam.
On Apr 18, 10:51 pm, Richard Norman <r_s_nor...@comcast.net> wrote:
> On Thu, 18 Apr 2013 18:37:50 -0700 (PDT), pnyikos
According to the Wikipedia article you cite below, the word "clade"
was already used for the modern phylogenetic concept by Julian Huxley
in 1958. Another Wikipedia entry attributes the word to him.
http://en.wikipedia.org/wiki/Clade
As for the concept, it would be a sorry systematist or even
paleontologist in the 20th who did not have it firmly in his mental
apparatus. Back before Hennig decided to change the usage of the word
"monophyletic taxon" to mean "clade," it was applied to both clades
and paraphyletic taxa.
>The introduction of
> purely (or relatively so) cladistic thinking within biology did not
> become generally accepted until the '70s or even later. Many of the
> groupings by pre-cladist biologicsts were based on possessing some
> shared characteristic, especially "level of organization" which is, in
> fact, exactly how a "grade" is defined.
Probably by Romer, see the comments above from an earlier post about
"subholostean" and "holostean".
In Colbert's _Evolution of the Vertebrates_, the word is used
differently ("family" is a grade, "order" is a grade, "class" is a
grade, etc. the way Colbert uses the word) but he does do the service
of listing the characters that make up the holostean grade,
contrasting them with the corresponding characters of chondrosteans on
the one hand and teleosts on the other.
But Harshman has a third idea about the meaning of "grade", again
wrapped up in characters rather than phylogeny.
> Wikipedia: In alpha taxonomy,
> a grade refers to a taxon united by a level of morphological or
> physiological complexity. The term was coined by British biologist
> Julian Huxley, to contrast with clade, a strictly phylogenetic unit.
> http://en.wikipedia.org/wiki/Evolutionary_grade
Wikipedia is not the final word on things like these. I'd rather
thrash these things out with Harshman, who at least has my attention
on this thread.
> Note: Huxley used the concept of clade somehwat less than a decade
> before Hennig but it didn't really catch on until people started
> arguing about Hennig's ideas.
>
> That source goes on to say: "Thus, traditional palaeontological works
> are often using evolutionary grades as formal or informal taxa,
> including examples such as Labyrinthodonts, Anapsids, Synapsids,
> Dinosaurs,
By the way, one of the reasons the taxon "Dinosauria" was resisted for
so long is that Romer had the Saurischians and Ornithiscians arising
independently from the Thecodonts, so Dinosauria would have been
polyphiletic.
> Ammonites, Eurypterid,
Huh? Don't ammonites form a clade? don't Eurypterids form one?
[Granted, there was a rather farfetched hypothesis that vertebrates
arose from Eurypterids that had been flipped over, but it never was
taken seriously by most, perhaps not even by its proposer(s).]
> Lobopodes and many of the more well
> known taxa of human evolution. Organizing organisms into grades rather
> than strict clades can also be very useful to understand the
> evolutionary sequence behind major diversification of both animals[6]
> and plants."
Yes.
> Interestingly, the Wikipedia page on cladistics uses, as an example,
> the disputed placement of turtles within and outsied the diapsids.
> http://en.wikipedia.org/wiki/Cladistics
>
> Further, I really don't understand why you argue so intensely about
> what Romer might have written in an era when molecular genetic data
> about phylogenetic relationships was essentially completely absent.
I argue intensely because of the way Romer has been misunderstood, and
with him the traditional systematists. The Wikipedia article claims
an equation grade = paraphyletic taxon, and that simply does not do
justice to Romer's thinking about paraphyletic taxa, not even by what
Harshman calls "the modern concept of grade."
Peter Nyikos
just in case someone interested in systematics chances across this
thread before it becomes buried under spam.
On Apr 18, 10:51 pm, Richard Norman <r_s_nor...@comcast.net> wrote:
> On Thu, 18 Apr 2013 18:37:50 -0700 (PDT), pnyikos
was already used for the modern phylogenetic concept by Julian Huxley
in 1958. Another Wikipedia entry attributes the word to him.
http://en.wikipedia.org/wiki/Clade
As for the concept, it would be a sorry systematist or even
paleontologist in the 20th who did not have it firmly in his mental
apparatus. Back before Hennig decided to change the usage of the word
"monophyletic taxon" to mean "clade," it was applied to both clades
and paraphyletic taxa.
>The introduction of
> purely (or relatively so) cladistic thinking within biology did not
> become generally accepted until the '70s or even later. Many of the
> groupings by pre-cladist biologicsts were based on possessing some
> shared characteristic, especially "level of organization" which is, in
> fact, exactly how a "grade" is defined.
"subholostean" and "holostean".
In Colbert's _Evolution of the Vertebrates_, the word is used
differently ("family" is a grade, "order" is a grade, "class" is a
grade, etc. the way Colbert uses the word) but he does do the service
of listing the characters that make up the holostean grade,
contrasting them with the corresponding characters of chondrosteans on
the one hand and teleosts on the other.
But Harshman has a third idea about the meaning of "grade", again
wrapped up in characters rather than phylogeny.
> Wikipedia: In alpha taxonomy,
> a grade refers to a taxon united by a level of morphological or
> physiological complexity. The term was coined by British biologist
> Julian Huxley, to contrast with clade, a strictly phylogenetic unit.
> http://en.wikipedia.org/wiki/Evolutionary_grade
thrash these things out with Harshman, who at least has my attention
on this thread.
> Note: Huxley used the concept of clade somehwat less than a decade
> before Hennig but it didn't really catch on until people started
> arguing about Hennig's ideas.
>
> That source goes on to say: "Thus, traditional palaeontological works
> are often using evolutionary grades as formal or informal taxa,
> including examples such as Labyrinthodonts, Anapsids, Synapsids,
> Dinosaurs,
so long is that Romer had the Saurischians and Ornithiscians arising
independently from the Thecodonts, so Dinosauria would have been
polyphiletic.
> Ammonites, Eurypterid,
Huh? Don't ammonites form a clade? don't Eurypterids form one?
[Granted, there was a rather farfetched hypothesis that vertebrates
arose from Eurypterids that had been flipped over, but it never was
taken seriously by most, perhaps not even by its proposer(s).]
> Lobopodes and many of the more well
> known taxa of human evolution. Organizing organisms into grades rather
> than strict clades can also be very useful to understand the
> evolutionary sequence behind major diversification of both animals[6]
> and plants."
> Interestingly, the Wikipedia page on cladistics uses, as an example,
> the disputed placement of turtles within and outsied the diapsids.
> http://en.wikipedia.org/wiki/Cladistics
>
> Further, I really don't understand why you argue so intensely about
> what Romer might have written in an era when molecular genetic data
> about phylogenetic relationships was essentially completely absent.
with him the traditional systematists. The Wikipedia article claims
an equation grade = paraphyletic taxon, and that simply does not do
justice to Romer's thinking about paraphyletic taxa, not even by what
Harshman calls "the modern concept of grade."
Peter Nyikos
pnyikos
Apr 22, 2013, 1:55:04 PM4/22/13
to nyi...@bellsouth.net
On Apr 19, 10:44 am, John Harshman <jharsh...@pacbell.net> wrote:
> On 4/18/13 6:37 PM, pnyikos wrote:
> > On Apr 18, 9:51 am, John Harshman<jharsh...@pacbell.net> wrote:
> >> On 4/17/13 8:57 PM, pnyikos wrote:
>
> >>> On Apr 17, 1:02 am, John Harshman<jharsh...@pacbell.net> wrote:
> >>>> On 4/16/13 7:54 PM, pnyikos wrote:
> >>>>> On Apr 13, 8:43 pm, John Harshman<jharsh...@pacbell.net> wrote:
> >>>>>> On 4/12/13 6:58 PM, pnyikos wrote:
>
> >>>>>>> On Apr 10, 6:51 pm, John Harshman<jharsh...@pacbell.net> wrote:
> On 4/18/13 6:37 PM, pnyikos wrote:
> > On Apr 18, 9:51 am, John Harshman<jharsh...@pacbell.net> wrote:
> >> On 4/17/13 8:57 PM, pnyikos wrote:
>
> >>> On Apr 17, 1:02 am, John Harshman<jharsh...@pacbell.net> wrote:
> >>>> On 4/16/13 7:54 PM, pnyikos wrote:
> >>>>> On Apr 13, 8:43 pm, John Harshman<jharsh...@pacbell.net> wrote:
> >>>>>> On 4/12/13 6:58 PM, pnyikos wrote:
>
> >>>>>>> On Apr 10, 6:51 pm, John Harshman<jharsh...@pacbell.net> wrote:
> >>>>>>>>>> That wasn't quite the "traditional" that Peter was talking about. He was
> >>>>>>>>>> referring, essentially, to gradism.
>
> >>>>>>>>> If you are referring to grades as in "subholostean grade" and
> >>>>>>>>> "holostean grade" you are dead wrong. That is the only example of
> >>>>>>>>> polyphyletic taxon that Romer gave in _Vertebrate Paleontology_, the
> >>>>>>>>> example of both the bowfin and the gar being in Holostei.
>
> >>>>>>> No reply to this from you, so I am still confused by what you mean by
> >>>>>>> "gradism." For now I will assume you really meant to refer to what I
> >>>>>>> described next:
>
> >>>>>> Not at all. Most grades are paraphyletic. "Reptile" is a grade.
> >>>>>> "Thecodont" is a grade. "Mammal-like reptile" is a grade. Traditional
> >>>>>> "Amphibia" is a grade.
>
> >>>>> All this is new to me. All the examples you name are simply
> >>>>> paraphyletic taxa to me, no more, no less.
>
> >>>> paraphyletic taxon = grade, mostly.
>
> >>> Sorry, I don't buy it.
>
> >> I can't help that.
>
> > You might if you were to tell me who the authority is on what "grades"
> > should mean to us, and where you get the idea that they were
> > extensively used by 20th century paleontologists before cladistics
> > came along.
>
> Perhaps it would be simpler is we were to stop fixating on the word
> "grade" and substitute "paraphyletic group", which means the same thing.
I see no such equivalence. "Paraphyletic group" is a purely
> >>>>>>>>>> referring, essentially, to gradism.
>
> >>>>>>>>> If you are referring to grades as in "subholostean grade" and
> >>>>>>>>> "holostean grade" you are dead wrong. That is the only example of
> >>>>>>>>> polyphyletic taxon that Romer gave in _Vertebrate Paleontology_, the
> >>>>>>>>> example of both the bowfin and the gar being in Holostei.
>
> >>>>>>> No reply to this from you, so I am still confused by what you mean by
> >>>>>>> "gradism." For now I will assume you really meant to refer to what I
> >>>>>>> described next:
>
> >>>>>> Not at all. Most grades are paraphyletic. "Reptile" is a grade.
> >>>>>> "Thecodont" is a grade. "Mammal-like reptile" is a grade. Traditional
> >>>>>> "Amphibia" is a grade.
>
> >>>>> All this is new to me. All the examples you name are simply
> >>>>> paraphyletic taxa to me, no more, no less.
>
> >>>> paraphyletic taxon = grade, mostly.
>
> >>> Sorry, I don't buy it.
>
> >> I can't help that.
>
> > You might if you were to tell me who the authority is on what "grades"
> > should mean to us, and where you get the idea that they were
> > extensively used by 20th century paleontologists before cladistics
> > came along.
>
> Perhaps it would be simpler is we were to stop fixating on the word
> "grade" and substitute "paraphyletic group", which means the same thing.
phylogenetic concept, as is "clade." You seem to have definite ideas
about the modern concept of "grade" but the description you've been
giving is too general to be of use and potentially misleading.
> Nor do I want to argue about the meaning of "grade", since it's
> peripheral to that point. (Though that's exactly what "grade" does mean.)
negating a great deal of what you have written up to now.
character-laden description of "the modern definition of grade"
suggests.
> >> which in
> >> fact may have been coined after he wrote. Nor is his understanding of
> >> the word "grade" -- or anything else -- necessarily the one we should
> >> take as our standard. I'm not arguing about Romer's definition. I'm
> >> arguing about whether the groups he liked fit our modern definition of
> >> "grade".
>
> > And who is the authority on what that definition is? Ferris, who was
> > under the delusion that all paraphyletic taxa are produced from a
> > clade by removing ONE clade when he had the example of Reptilia, with
> > both Aves and Mammalia removed, to tell him otherwise?
> > I do believe, though, that he was more on the right track than whoever
> > it was that came up with the formula that "monophyletic taxa are
> > defined by apomorphies, paraphyletic by plesimorphies, and
> > polyphyletic by similarity."
> We
> aren't arguing about the definition of paraphyly. And I'm not sure who
> "Ferris" is, unless you refer to Steve Farris.
> Anyway, we all here agree
> on what paraphyly is.
also Farris who came up with that slogan "...parapyletic taxa are
defined by plesimorphies..."?
Continued in next post.
Peter Nyikos
pnyikos
Apr 22, 2013, 2:16:23 PM4/22/13
to nyi...@bellsouth.net
On Apr 19, 10:44 am, John Harshman <jharsh...@pacbell.net> wrote:
> On 4/18/13 6:37 PM, pnyikos wrote:
> > On Apr 18, 9:51 am, John Harshman<jharsh...@pacbell.net> wrote:
> >> On 4/17/13 8:57 PM, pnyikos wrote:
>
> >>> On Apr 17, 1:02 am, John Harshman<jharsh...@pacbell.net> wrote:
> >>>> On 4/16/13 7:54 PM, pnyikos wrote:
> >>>>> On Apr 13, 8:43 pm, John Harshman<jharsh...@pacbell.net> wrote:
> >>>>>> On 4/12/13 6:58 PM, pnyikos wrote:
> > Romer et. al. produced their
> > "paraphyletic" taxa via "horizontal cleavage" or else they inherited
> > them from earlier systematists, threw out the ones that were
> > polyphyletic (with a few exceptions like Holostei, which were looked
> > upon as "necessary evils") and hung on to the "true phyletic groups,"
> > both paraphyletic and holophyletic.
>
> Again, we agree that Romer liked paraphyletic taxa.
Yes, in the phylogenetic sense; see my previous reply. By the way,
> On 4/18/13 6:37 PM, pnyikos wrote:
> > On Apr 18, 9:51 am, John Harshman<jharsh...@pacbell.net> wrote:
> >> On 4/17/13 8:57 PM, pnyikos wrote:
>
> >>> On Apr 17, 1:02 am, John Harshman<jharsh...@pacbell.net> wrote:
> >>>> On 4/16/13 7:54 PM, pnyikos wrote:
> >>>>> On Apr 13, 8:43 pm, John Harshman<jharsh...@pacbell.net> wrote:
> >>>>>> On 4/12/13 6:58 PM, pnyikos wrote:
> > Romer et. al. produced their
> > "paraphyletic" taxa via "horizontal cleavage" or else they inherited
> > them from earlier systematists, threw out the ones that were
> > polyphyletic (with a few exceptions like Holostei, which were looked
> > upon as "necessary evils") and hung on to the "true phyletic groups,"
> > both paraphyletic and holophyletic.
>
> Again, we agree that Romer liked paraphyletic taxa.
just so there is no misunderstanding: by "phylogenetic" I mean "based
on the best current estimate of the topology of the Tree of Life."
Accordingly, I consider the term "phylogenetic systematics" to be a
misleading expression for "classifications that avoid all paraphyletic
taxa".
[snip]
> >>>> I believe you misunderstand the meaning of "grade". No measure of value
> >>>> or sophistication is intended. Grades are merely collections of species
> >>>> that resemble each other in particular characters. When those characters
> >>>> do not define clades, the grade is paraphyletic.
>
> >>> False, see the word "polyphyletic" up there. And I would expect the
> >>> old traditional paleontology texts to be littered with polyphyletic
> >>> "taxa" if things were as you describe. They aren't.
>
> >>>> As in "skin covered by
> >>>> keratinous scales" defines a grade Reptilia.
>
> >>> Except when it doesn't, as in fossils where there is no sign of
> >>> keratin. Then Romer would have relied on<gasp> phylogeny.
>
> >> Which he would use to define a paraphyletic group, Reptilia. And that
> >> group would be based on the presence of certain character states,
>
> > as are clades.
>
> You are wrong about that. Clades are based on phylogeny and nothing
> more.
aren't clades determined by a single apomorphy apiece? That's how
cladists produce phylogenetic trees, isn't it?
> While phylogeny is estimated using characters, a clade is not
> based on particular character states, and in fact there may be no state
> universal within a clade.
covered scales". These could have been "lost" many times within the
old Reptila, but it was their "loss" [= replacement by feathers] that
cut Aves away from them. And it was mostly a matter of "liberating"
Aves rather than deciding that the modification of one character had
gone just too far.
Mammalia may be a similar case, hair in place of feathers, but for the
paleontologists I think the real dividing line had to do with the
movement of two tiny bones in the hinge of the jaw migrating into the
middle ear. "lower jaw = dentary" was the way it was often put.
> >> and
> >> would exclude those with more derived states. That's what a grade is.
>
> > Romer's practice of systematics does not fit that Procrustean bed.
> > See above.
>
> And above you show a number of paraphyletic groups, formed by excluding
> taxa with derived states.
cases.
One is where the discovery of Archaeopteryx was probably instrumental
in the decision to use feathers to mark the transition between the
classes "Reptilia" and "Aves" which had been inherited from Linne
(Linnaeus).
It wasn't until about a century later that feathers may have turned
out to be the wrong place to put the dividing line. Had the
systematists of old seen lots of fossils of feathered dinosaurs that
obviously were *at best* in the "proavis" stage, they might have put
the dividing line at "Archie" anyway, but would have given different
reasons for putting the dividing line there.
By the way, I saw pictures back in 1956 of two versions of the
hypothetical "Proavis," the tree-climber version and the running
version, and although each had feathers, the question of whether to
class them as birds or reptiles was not addressed IIRC. Perhaps the
situation was like the on-again-off-again case of whether to classify
*Bienotherium* as a mammal or as a "mammal-like reptile."
Peter Nyikos
Richard Norman
Apr 22, 2013, 6:01:05 PM4/22/13
to
violent revolution in thinking about systematics in the 1970's period.
Yes, it is true that J Huxley did propose the term "clade" before
Hennig, as I clearly indicated later in my post. I also indicated
that the term did not catch on as "the" important organizing concept
until the Hennigists took over. With cladistics, paraphyletic groups
are simply improper groups -- they are wrong. Before that,
paleontologists quite happily accepted them.
You seem to be hung up on ensuring that technical terms have single,
unambiguous, and unchanging definitions. That some "old" or
"traditional" grades are monophyletic while others are polyphyletic
really doesn't matter. The point is whether or not the grouping is
defined by modern phologenetic systematics ("cladistics").
If you doubt the "violence" (yes I deliberately chose a very loaded
term) of the cladistics revolution, look at criticisms such as
http://www.kheper.net/evolution/systematics/cladistics.htm
and look at Cracraft's contemporary (1981) paper essentially urging
palontologists to accept cladistics and outlining the reaction against
it
http://research.amnh.org/vz/ornithology/pdfs/1981e.Pattern%20process%20in%20paleobiology.pdf
John Harshman
Apr 22, 2013, 6:18:29 PM4/22/13
to
paraphyletic groups is because they share characters that are
transformed in the groups detached from them. And that's a grade. So in
practice there is no distinction. It would certainly be possible to
define paraphyletic groups that weren't grades (amniota minus Carnivora,
for example), but who would do that?
>> Nor do I want to argue about the meaning of "grade", since it's
>> peripheral to that point. (Though that's exactly what "grade" does mean.)
>
> Then I suggest you drop all talk about characters, even at the cost of
> negating a great deal of what you have written up to now.
>>>> which in
>>>> fact may have been coined after he wrote. Nor is his understanding of
>>>> the word "grade" -- or anything else -- necessarily the one we should
>>>> take as our standard. I'm not arguing about Romer's definition. I'm
>>>> arguing about whether the groups he liked fit our modern definition of
>>>> "grade".
>>
>>> And who is the authority on what that definition is? Ferris, who was
>>> under the delusion that all paraphyletic taxa are produced from a
>>> clade by removing ONE clade when he had the example of Reptilia, with
>>> both Aves and Mammalia removed, to tell him otherwise?
>>> I do believe, though, that he was more on the right track than whoever
>>> it was that came up with the formula that "monophyletic taxa are
>>> defined by apomorphies, paraphyletic by plesimorphies, and
>>> polyphyletic by similarity."
>
> [snip snarky aside by you]
>
>> We
>> aren't arguing about the definition of paraphyly. And I'm not sure who
>> "Ferris" is, unless you refer to Steve Farris.
>
> Yes. My apologies for the misspelling.
>
>> Anyway, we all here agree
>> on what paraphyly is.
>
> ...in the strictly phylogenetic sense, yes. By the way, wasn't it
> also Farris who came up with that slogan "...parapyletic taxa are
> defined by plesimorphies..."?
John Harshman
Apr 22, 2013, 6:23:40 PM4/22/13
to
On 4/22/13 11:16 AM, pnyikos wrote:
> On Apr 19, 10:44 am, John Harshman<jharsh...@pacbell.net> wrote:
>> On 4/18/13 6:37 PM, pnyikos wrote:
>
>>> On Apr 18, 9:51 am, John Harshman<jharsh...@pacbell.net> wrote:
>>>> On 4/17/13 8:57 PM, pnyikos wrote:
>>
>>>>> On Apr 17, 1:02 am, John Harshman<jharsh...@pacbell.net> wrote:
>>>>>> On 4/16/13 7:54 PM, pnyikos wrote:
>>>>>>> On Apr 13, 8:43 pm, John Harshman<jharsh...@pacbell.net> wrote:
>>>>>>>> On 4/12/13 6:58 PM, pnyikos wrote:
>
>
>>> Romer et. al. produced their
>>> "paraphyletic" taxa via "horizontal cleavage" or else they inherited
>>> them from earlier systematists, threw out the ones that were
>>> polyphyletic (with a few exceptions like Holostei, which were looked
>>> upon as "necessary evils") and hung on to the "true phyletic groups,"
>>> both paraphyletic and holophyletic.
>>
>> Again, we agree that Romer liked paraphyletic taxa.
>
> Yes, in the phylogenetic sense; see my previous reply. By the way,
> just so there is no misunderstanding: by "phylogenetic" I mean "based
> on the best current estimate of the topology of the Tree of Life."
>
> Accordingly, I consider the term "phylogenetic systematics" to be a
> misleading expression for "classifications that avoid all paraphyletic
> taxa".
Nobody cares as long as we all know what it means. How do you like
> On Apr 19, 10:44 am, John Harshman<jharsh...@pacbell.net> wrote:
>> On 4/18/13 6:37 PM, pnyikos wrote:
>
>>> On Apr 18, 9:51 am, John Harshman<jharsh...@pacbell.net> wrote:
>>>> On 4/17/13 8:57 PM, pnyikos wrote:
>>
>>>>> On Apr 17, 1:02 am, John Harshman<jharsh...@pacbell.net> wrote:
>>>>>> On 4/16/13 7:54 PM, pnyikos wrote:
>>>>>>> On Apr 13, 8:43 pm, John Harshman<jharsh...@pacbell.net> wrote:
>>>>>>>> On 4/12/13 6:58 PM, pnyikos wrote:
>
>
>>> Romer et. al. produced their
>>> "paraphyletic" taxa via "horizontal cleavage" or else they inherited
>>> them from earlier systematists, threw out the ones that were
>>> polyphyletic (with a few exceptions like Holostei, which were looked
>>> upon as "necessary evils") and hung on to the "true phyletic groups,"
>>> both paraphyletic and holophyletic.
>>
>> Again, we agree that Romer liked paraphyletic taxa.
>
> Yes, in the phylogenetic sense; see my previous reply. By the way,
> just so there is no misunderstanding: by "phylogenetic" I mean "based
> on the best current estimate of the topology of the Tree of Life."
>
> Accordingly, I consider the term "phylogenetic systematics" to be a
> misleading expression for "classifications that avoid all paraphyletic
> taxa".
Mayr's "evolutionary systematics"?
>>>>>> I believe you misunderstand the meaning of "grade". No measure of value
>>>>>> or sophistication is intended. Grades are merely collections of species
>>>>>> that resemble each other in particular characters. When those characters
>>>>>> do not define clades, the grade is paraphyletic.
>>
>>>>> False, see the word "polyphyletic" up there. And I would expect the
>>>>> old traditional paleontology texts to be littered with polyphyletic
>>>>> "taxa" if things were as you describe. They aren't.
>>
>>>>>> As in "skin covered by
>>>>>> keratinous scales" defines a grade Reptilia.
>>
>>>>> Except when it doesn't, as in fossils where there is no sign of
>>>>> keratin. Then Romer would have relied on<gasp> phylogeny.
>>
>>>> Which he would use to define a paraphyletic group, Reptilia. And that
>>>> group would be based on the presence of certain character states,
>>
>>> as are clades.
>>
>> You are wrong about that. Clades are based on phylogeny and nothing
>> more.
>
> Nice to see such a strong statement by you. But in actual practice,
> aren't clades determined by a single apomorphy apiece? That's how
> cladists produce phylogenetic trees, isn't it?
have many, or in fact it may have none, at least none that are universal
to all extant members.
>> While phylogeny is estimated using characters, a clade is not
>> based on particular character states, and in fact there may be no state
>> universal within a clade.
>
> The same is true of paraphyletic taxa. Take your example of "keratin
> covered scales". These could have been "lost" many times within the
> old Reptila, but it was their "loss" [= replacement by feathers] that
> cut Aves away from them. And it was mostly a matter of "liberating"
> Aves rather than deciding that the modification of one character had
> gone just too far.
>
> Mammalia may be a similar case, hair in place of feathers, but for the
> paleontologists I think the real dividing line had to do with the
> movement of two tiny bones in the hinge of the jaw migrating into the
> middle ear. "lower jaw = dentary" was the way it was often put.
>>>> and
>>>> would exclude those with more derived states. That's what a grade is.
>>
>>> Romer's practice of systematics does not fit that Procrustean bed.
>>> See above.
>>
>> And above you show a number of paraphyletic groups, formed by excluding
>> taxa with derived states.
>
> "with derived states" is putting the cart before the horse in some
> cases.
do that.
> One is where the discovery of Archaeopteryx was probably instrumental
> in the decision to use feathers to mark the transition between the
> classes "Reptilia" and "Aves" which had been inherited from Linne
> (Linnaeus).
sharper. I do not see your point.
pnyikos
Apr 24, 2013, 12:37:22 PM4/24/13
to nyi...@bellsouth.net
> >> Note: Huxley used the concept of clade somehwat less than a decade
> >> before Hennig but it didn't really catch on until people started
> >> arguing about Hennig's ideas.
>
> >> That source goes on to say: "Thus, traditional palaeontological works
> >> are often using evolutionary grades as formal or informal taxa,
> >> including examples such as Labyrinthodonts, Anapsids, Synapsids,
> >> Dinosaurs,
>
> >By the way, one of the reasons the taxon "Dinosauria" was resisted for
> >so long is that Romer had the Saurischians and Ornithiscians arising
> >independently from the Thecodonts, so Dinosauria would have been
> >polyphiletic.
>>> Organizing organisms into grades rather
> >> than strict clades can also be very useful to understand the
> >> evolutionary sequence behind major diversification of both animals[6]
> >> and plants."
>
> >Yes.
>
> >> Interestingly, the Wikipedia page on cladistics uses, as an example,
> >> the disputed placement of turtles within and outsied the diapsids.
> >> http://en.wikipedia.org/wiki/Cladistics
>
> >> Further, I really don't understand why you argue so intensely about
> >> what Romer might have written in an era when molecular genetic data
> >> about phylogenetic relationships was essentially completely absent.
>
> >I argue intensely because of the way Romer has been misunderstood, and
> >with him the traditional systematists. The Wikipedia article claims
> >an equation grade = paraphyletic taxon, and that simply does not do
> >justice to Romer's thinking about paraphyletic taxa, not even by what
> >Harshman calls "the modern concept of grade."
>
> I think what you completely fail to understand is there was a rather
> violent revolution in thinking about systematics in the 1970's period.
Actually, I've read about it in a book by one of the chief "rebels".
> >> than strict clades can also be very useful to understand the
> >> evolutionary sequence behind major diversification of both animals[6]
> >> and plants."
>
> >Yes.
>
> >> Interestingly, the Wikipedia page on cladistics uses, as an example,
> >> the disputed placement of turtles within and outsied the diapsids.
> >> http://en.wikipedia.org/wiki/Cladistics
>
> >> Further, I really don't understand why you argue so intensely about
> >> what Romer might have written in an era when molecular genetic data
> >> about phylogenetic relationships was essentially completely absent.
>
> >I argue intensely because of the way Romer has been misunderstood, and
> >with him the traditional systematists. The Wikipedia article claims
> >an equation grade = paraphyletic taxon, and that simply does not do
> >justice to Romer's thinking about paraphyletic taxa, not even by what
> >Harshman calls "the modern concept of grade."
>
> I think what you completely fail to understand is there was a rather
> violent revolution in thinking about systematics in the 1970's period.
But he was so ignorant of the traditional meaning of "related" that
he misunderstood a complaint by one of the "old guard" that now
lungfish will be considered to be more closely related to horses than
to salmon.
He actually thought that this old-timer was claiming that lungfish and
salmon were in a clade which excluded horses, and drew a tree
accordingly to show how the old-timer was ignorant of the tree of
life.
> Yes, it is true that J Huxley did propose the term "clade" before
> Hennig, as I clearly indicated later in my post. I also indicated
> that the term did not catch on as "the" important organizing concept
> until the Hennigists took over. With cladistics, paraphyletic groups
> are simply improper groups -- they are wrong.
Look, I once floated the idea of reforming the Linnean classification
so that it, too, could reflect hypothesized phylogeny just as well as
cladistic classification. It involved putting certain key organisms
into two, partially overlapping taxa. Example: Archaeopteryx would be
in *both* the class Dinosauria and the class Aves, with the overlap
being the infraorder Maniraptora.
It was rejected by the people in sci.bio.paleontology and
sci.bio.systematics [both of which were healthy going concerns at the
time] as being too compllicated. But I'll be glad to show it to you
if you are interested.
> Before that,
> paleontologists quite happily accepted them.
their number diminish.
However, some newcomers are beginning to feel that there is something
missing in their lives. When I have a little more time, I'll try to
find an url in my voluminous files that will take you to one of them.
> You seem to be hung up on ensuring that technical terms have single,
> unambiguous, and unchanging definitions.
the underlying issues.
And now I must run. Duty calls. Thanks for the urls, I'll gladly
take a look when I have more time to spare from grading.
Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer--
University of South Carolina
http://www.math.sc.edu/~nyikos/
nyikos @ math.sc.edu
> That some "old" or
> "traditional" grades are monophyletic while others are polyphyletic
> really doesn't matter. The point is whether or not the grouping is
> defined by modern phologenetic systematics ("cladistics").
>
> If you doubt the "violence" (yes I deliberately chose a very loaded
> term) of the cladistics revolution, look at criticisms such as
> http://www.kheper.net/evolution/systematics/cladistics.htm
> and look at Cracraft's contemporary (1981) paper essentially urging
> palontologists to accept cladistics and outlining the reaction against
> it
>
> http://research.amnh.org/vz/ornithology/pdfs/1981e.Pattern%20process%...
> "traditional" grades are monophyletic while others are polyphyletic
> really doesn't matter. The point is whether or not the grouping is
> defined by modern phologenetic systematics ("cladistics").
>
> If you doubt the "violence" (yes I deliberately chose a very loaded
> term) of the cladistics revolution, look at criticisms such as
> http://www.kheper.net/evolution/systematics/cladistics.htm
> and look at Cracraft's contemporary (1981) paper essentially urging
> palontologists to accept cladistics and outlining the reaction against
> it
>
Richard Norman
Apr 24, 2013, 2:53:47 PM4/24/13
to
On Wed, 24 Apr 2013 09:37:22 -0700 (PDT), pnyikos
<nyi...@bellsouth.net> wrote:
>On Apr 22, 6:01 pm, Richard Norman <r_s_nor...@comcast.net> wrote:
>> On Mon, 22 Apr 2013 10:30:32 -0700 (PDT), pnyikos
<snip to eliminate ancient history>
<nyi...@bellsouth.net> wrote:
>On Apr 22, 6:01 pm, Richard Norman <r_s_nor...@comcast.net> wrote:
>> On Mon, 22 Apr 2013 10:30:32 -0700 (PDT), pnyikos
>>
>> I think what you completely fail to understand is there was a rather
>> violent revolution in thinking about systematics in the 1970's period.
>
>Actually, I've read about it in a book by one of the chief "rebels".
>But he was so ignorant of the traditional meaning of "related" that
>he misunderstood a complaint by one of the "old guard" that now
>lungfish will be considered to be more closely related to horses than
>to salmon.
>
>He actually thought that this old-timer was claiming that lungfish and
>salmon were in a clade which excluded horses, and drew a tree
>accordingly to show how the old-timer was ignorant of the tree of
>life.
written as a misunderstanding of what was held previously. But the
fact is that there really was a revolution in thinking. You seem not
to like it but it is a fait accompli and has been for at least a
quarter century.
>> Yes, it is true that J Huxley did propose the term "clade" before
>> Hennig, as I clearly indicated later in my post. I also indicated
>> that the term did not catch on as "the" important organizing concept
>> until the Hennigists took over. With cladistics, paraphyletic groups
>> are simply improper groups -- they are wrong.
>
>They are eminently right for the purposes to which they are put.
>
>Look, I once floated the idea of reforming the Linnean classification
>so that it, too, could reflect hypothesized phylogeny just as well as
>cladistic classification. It involved putting certain key organisms
>into two, partially overlapping taxa. Example: Archaeopteryx would be
>in *both* the class Dinosauria and the class Aves, with the overlap
>being the infraorder Maniraptora.
>
>It was rejected by the people in sci.bio.paleontology and
>sci.bio.systematics [both of which were healthy going concerns at the
>time] as being too compllicated. But I'll be glad to show it to you
>if you are interested.
However the field of systematics has proceeded quite nicely after that
"revolution" with the cladistic way of thinking. Paraphyletic groups
are convenient for a number of reasons -- herpetologists and
invertebrate biologists and icthyologists are still in business -- but
they are rightly rejected for a formal analysis of phylogenetics.
If you have a superior system you will have to get it published in a
systematics journal and then have the mass of systemecists acknowledge
its superiority and switch. Showing it to two news groups won't work
but is a good test as a trial balloon. You did try it and it failed.
I can understand your frustration and even anger that your idea was
not accepted but unfortunately that is the way the world works.
>> Before that,
>> paleontologists quite happily accepted them.
>
>And many of the old-timers clung to them. Harshman rejoices to see
>their number diminish.
>
>However, some newcomers are beginning to feel that there is something
>missing in their lives. When I have a little more time, I'll try to
>find an url in my voluminous files that will take you to one of them.
>
(actually now two generations) are somewhat impatient about the
holdovers. And it is to be expected that there are upstarts unhappy
with the current situation and eager to press on new ways of thinking.
That is the way it works. Most of them will be rejected just as your
proposal was. Maybe one of them will be the new Hennig.
pnyikos
Apr 25, 2013, 4:00:51 PM4/25/13
to nyi...@bellsouth.net
Yesterday Richard Norman got my one reply on this thread, and today it
is you, John. I'm almost as busy today as I was yesterday, and no
relief is in sight until next week.
Since it is out of style, I don't see any point in talking about it.
> >>>>>> I believe you misunderstand the meaning of "grade". No measure of value
> >>>>>> or sophistication is intended. Grades are merely collections of species
> >>>>>> that resemble each other in particular characters. When those characters
> >>>>>> do not define clades, the grade is paraphyletic.
>
> >>>>> False, see the word "polyphyletic" up there. �And I would expect the
> >>>>> old traditional paleontology texts to be littered with polyphyletic
> >>>>> "taxa" if things were as you describe. They aren't.
>
> >>>>>> As in "skin covered by
> >>>>>> keratinous scales" defines a grade Reptilia.
>
> >>>>> Except when it doesn't, as in fossils where there is no sign of
> >>>>> keratin. �Then Romer would have relied on<gasp> � � phylogeny.
>
> >>>> Which he would use to define a paraphyletic group, Reptilia. And that
> >>>> group would be based on the presence of certain character states,
>
> >>> as are clades.
>
> >> You are wrong about that. Clades are based on phylogeny and nothing
> >> more.
>
> > Nice to see such a strong statement by you. But in actual practice,
> > aren't clades determined by a single apomorphy apiece? �That's how
> > cladists produce phylogenetic trees, isn't it?
>
> No, it isn't. It may be that a clade has a single apomorphy. Or it may
> have many, or in fact it may have none, at least none that are universal
> to all extant members.
Thanks. I presume that, as far as fossilizable materials go, this is
because the fossil record is so fragmentary that when one looks at the
characters of something like "Archie", a considerable number of
apomorphies may have originated in various ancestral "ghost species."
> >> While phylogeny is estimated using characters, a clade is not
> >> based on particular character states, and in fact there may be no state
> >> universal within a clade.
>
> > The same is true of paraphyletic taxa. �Take your example of "keratin
> > covered scales". �These could have been "lost" many times within the
> > old Reptila, but it was their "loss" [= replacement by feathers] that
> > cut Aves away from them. �And it was mostly a matter of "liberating"
> > Aves rather than deciding that the modification of one character had
> > gone just too far.
>
> > Mammalia may be a similar case, hair in place of feathers, but for the
> > paleontologists I think the real dividing line had to do with the
> > movement of two tiny bones in the hinge of the jaw migrating into the
> > middle ear. �"lower jaw = dentary" was the way it was often put.
>
> I do not see the distinction you are trying to make.
The distinction is between vast numbers of "derived" states of a vast
number of characters. Only a very small, very select few apomorphies
are used to cut taxa like Aves and Mammalia away from taxa like old
Reptilia.
That is why it is so silly to say that Reptilia is determined by
plesimorphies, and why the word "derived" is only the bare bones of a
beginning of an explanation of how paraphyletic taxa are arrived at.
Just look at what appears next, of what you wrote back then:
> >>>> and
> >>>> would exclude those with more derived states. That's what a grade is.
That's the kind of talk that could lull students, graduate and
undergraduate, into thinking that they know what traditonal
systematics, the one supplanted by the "violent revolution of the
seventies," was all about.
Do ANY of them get any more of a grounding in the in the old ways, or
do they just have to pick it up on their own?
Taxa like Mammalia and Aves took on huge significance in the old
classification. The way things are going, we may be swamped with so
many clades containing, and contained in, these two, that students are
apt to be unable to see the forest for the trees.
> >>> Romer's practice of �systematics does not fit that Procrustean bed.
> >>> See above.
>
> >> And above you show a number of paraphyletic groups, formed by excluding
> >> taxa with derived states.
>
> > "with derived states" is putting the cart before the horse in some
> > cases.
>
> You will have to explain what you mean. And what's below doesn't seem to
> do that.
The cutoffs come first, the derived states that are then used to make
those cutoffs clear could come a lot later.
> > One is where the discovery of Archaeopteryx was probably instrumental
> > in the decision to use feathers to mark the transition between the
> > classes "Reptilia" and "Aves" which had been inherited from Linne
> > (Linnaeus).
>
> I doubt that. Without Archaeopteryx, in fact, the dividing line is much
> sharper.
You are just reinforcing what I wrote: when Archie was discovered,
some decisions had to be made as to which side of the line it fell on,
and how one is thenceforth to decide on which side of the line a new
fossil will go.
> I do not see your point.
Thanks for leaving in the rest of what I wrote below. Maybe a second
reading will give you a better idea of what my point was than the
first did, now that I've added some more comments.
Peter Nyikos
is you, John. I'm almost as busy today as I was yesterday, and no
relief is in sight until next week.
> >>>>>> I believe you misunderstand the meaning of "grade". No measure of value
> >>>>>> or sophistication is intended. Grades are merely collections of species
> >>>>>> that resemble each other in particular characters. When those characters
> >>>>>> do not define clades, the grade is paraphyletic.
>
> >>>>> False, see the word "polyphyletic" up there. �And I would expect the
> >>>>> old traditional paleontology texts to be littered with polyphyletic
> >>>>> "taxa" if things were as you describe. They aren't.
>
> >>>>>> As in "skin covered by
> >>>>>> keratinous scales" defines a grade Reptilia.
>
> >>>>> Except when it doesn't, as in fossils where there is no sign of
> >>>>> keratin. �Then Romer would have relied on<gasp> � � phylogeny.
>
> >>>> Which he would use to define a paraphyletic group, Reptilia. And that
> >>>> group would be based on the presence of certain character states,
>
> >>> as are clades.
>
> >> You are wrong about that. Clades are based on phylogeny and nothing
> >> more.
>
> > Nice to see such a strong statement by you. But in actual practice,
> > aren't clades determined by a single apomorphy apiece? �That's how
> > cladists produce phylogenetic trees, isn't it?
>
> No, it isn't. It may be that a clade has a single apomorphy. Or it may
> have many, or in fact it may have none, at least none that are universal
> to all extant members.
because the fossil record is so fragmentary that when one looks at the
characters of something like "Archie", a considerable number of
apomorphies may have originated in various ancestral "ghost species."
> >> While phylogeny is estimated using characters, a clade is not
> >> based on particular character states, and in fact there may be no state
> >> universal within a clade.
>
> > The same is true of paraphyletic taxa. �Take your example of "keratin
> > covered scales". �These could have been "lost" many times within the
> > old Reptila, but it was their "loss" [= replacement by feathers] that
> > cut Aves away from them. �And it was mostly a matter of "liberating"
> > Aves rather than deciding that the modification of one character had
> > gone just too far.
>
> > Mammalia may be a similar case, hair in place of feathers, but for the
> > paleontologists I think the real dividing line had to do with the
> > movement of two tiny bones in the hinge of the jaw migrating into the
> > middle ear. �"lower jaw = dentary" was the way it was often put.
>
> I do not see the distinction you are trying to make.
number of characters. Only a very small, very select few apomorphies
are used to cut taxa like Aves and Mammalia away from taxa like old
Reptilia.
That is why it is so silly to say that Reptilia is determined by
plesimorphies, and why the word "derived" is only the bare bones of a
beginning of an explanation of how paraphyletic taxa are arrived at.
Just look at what appears next, of what you wrote back then:
> >>>> and
> >>>> would exclude those with more derived states. That's what a grade is.
undergraduate, into thinking that they know what traditonal
systematics, the one supplanted by the "violent revolution of the
seventies," was all about.
Do ANY of them get any more of a grounding in the in the old ways, or
do they just have to pick it up on their own?
Taxa like Mammalia and Aves took on huge significance in the old
classification. The way things are going, we may be swamped with so
many clades containing, and contained in, these two, that students are
apt to be unable to see the forest for the trees.
> >>> Romer's practice of �systematics does not fit that Procrustean bed.
> >>> See above.
>
> >> And above you show a number of paraphyletic groups, formed by excluding
> >> taxa with derived states.
>
> > "with derived states" is putting the cart before the horse in some
> > cases.
>
> You will have to explain what you mean. And what's below doesn't seem to
> do that.
those cutoffs clear could come a lot later.
> > One is where the discovery of Archaeopteryx was probably instrumental
> > in the decision to use feathers to mark the transition between the
> > classes "Reptilia" and "Aves" which had been inherited from Linne
> > (Linnaeus).
>
> I doubt that. Without Archaeopteryx, in fact, the dividing line is much
> sharper.
some decisions had to be made as to which side of the line it fell on,
and how one is thenceforth to decide on which side of the line a new
fossil will go.
> I do not see your point.
reading will give you a better idea of what my point was than the
first did, now that I've added some more comments.
Peter Nyikos
John Harshman
Apr 25, 2013, 4:44:13 PM4/25/13
to
misleading names.
characters from various parts of the anatomy, that wasn't the point. The
point was that any character present at the root of a clade may be
transformed in any descendant, and it's possible that every character
has been transformed in at least one descendant, leaving the clade with
no universal synapomorphies.
>>>> While phylogeny is estimated using characters, a clade is not
>>>> based on particular character states, and in fact there may be no state
>>>> universal within a clade.
>>
>>> The same is true of paraphyletic taxa. Take your example of "keratin
>>> covered scales". These could have been "lost" many times within the
>>> old Reptila, but it was their "loss" [= replacement by feathers] that
>>> cut Aves away from them. And it was mostly a matter of "liberating"
>>> Aves rather than deciding that the modification of one character had
>>> gone just too far.
>>
>>> Mammalia may be a similar case, hair in place of feathers, but for the
>>> paleontologists I think the real dividing line had to do with the
>>> movement of two tiny bones in the hinge of the jaw migrating into the
>>> middle ear. "lower jaw = dentary" was the way it was often put.
>>
>> I do not see the distinction you are trying to make.
>
> The distinction is between vast numbers of "derived" states of a vast
> number of characters. Only a very small, very select few apomorphies
> are used to cut taxa like Aves and Mammalia away from taxa like old
> Reptilia.
rarely preserved soft anatomy, so paleontologists use different ones,
not because they thought they were more important but because they were
more likely to be preserved. And this points to 2) the apomorphies
chosen and the paraphyletic groups so defined are arbitrary and capricious.
> That is why it is so silly to say that Reptilia is determined by
> plesimorphies, and why the word "derived" is only the bare bones of a
> beginning of an explanation of how paraphyletic taxa are arrived at.
plesiomorphies, which you choose to think of as the logically equivalent
criterion "lack of apomorphies".
> Just look at what appears next, of what you wrote back then:
>
>>>>>> and
>>>>>> would exclude those with more derived states. That's what a grade is.
>
> That's the kind of talk that could lull students, graduate and
> undergraduate, into thinking that they know what traditonal
> systematics, the one supplanted by the "violent revolution of the
> seventies," was all about.
>
> Do ANY of them get any more of a grounding in the in the old ways, or
> do they just have to pick it up on their own?
> Taxa like Mammalia and Aves took on huge significance in the old
> classification. The way things are going, we may be swamped with so
> many clades containing, and contained in, these two, that students are
> apt to be unable to see the forest for the trees.
likelihood that there will be a problem.
>>>>> Romer's practice of systematics does not fit that Procrustean bed.
>>>>> See above.
>>
>>>> And above you show a number of paraphyletic groups, formed by excluding
>>>> taxa with derived states.
>>
>>> "with derived states" is putting the cart before the horse in some
>>> cases.
>>
>> You will have to explain what you mean. And what's below doesn't seem to
>> do that.
>
> The cutoffs come first, the derived states that are then used to make
> those cutoffs clear could come a lot later.
>>> One is where the discovery of Archaeopteryx was probably instrumental
>>> in the decision to use feathers to mark the transition between the
>>> classes "Reptilia" and "Aves" which had been inherited from Linne
>>> (Linnaeus).
>>
>> I doubt that. Without Archaeopteryx, in fact, the dividing line is much
>> sharper.
>
> You are just reinforcing what I wrote: when Archie was discovered,
> some decisions had to be made as to which side of the line it fell on,
> and how one is thenceforth to decide on which side of the line a new
> fossil will go.
longer has to do anything of the sort. What one has to do now is to
attempt an estimate of phylogeny, i.e. to decide what groups
Archaeopteryx and everything else fall into.
>> I do not see your point.
>
> Thanks for leaving in the rest of what I wrote below. Maybe a second
> reading will give you a better idea of what my point was than the
> first did, now that I've added some more comments.
Walter Bushell
Apr 26, 2013, 8:18:20 PM4/26/13
to
In article <c8udncWehc-...@giganews.com>,
that group.
--
Gambling with Other People's Money is the meth of the fiscal industry.
me -- in the spirit of Karl and Groucho Marx
John Harshman <jhar...@pacbell.net> wrote:
> Technically, you might be correct. But the only reason people erect
> paraphyletic groups is because they share characters that are
> transformed in the groups detached from them. And that's a grade. So in
> practice there is no distinction. It would certainly be possible to
> define paraphyletic groups that weren't grades (amniota minus Carnivora,
> for example), but who would do that?
You just did. In fact to understand you statement I had to contemplate
> Technically, you might be correct. But the only reason people erect
> paraphyletic groups is because they share characters that are
> transformed in the groups detached from them. And that's a grade. So in
> practice there is no distinction. It would certainly be possible to
> define paraphyletic groups that weren't grades (amniota minus Carnivora,
> for example), but who would do that?
that group.
--
Gambling with Other People's Money is the meth of the fiscal industry.
me -- in the spirit of Karl and Groucho Marx
John Harshman
Apr 26, 2013, 8:38:58 PM4/26/13
to
On 4/26/13 5:18 PM, Walter Bushell wrote:
> In article<c8udncWehc-...@giganews.com>,
> John Harshman<jhar...@pacbell.net> wrote:
>
>> Technically, you might be correct. But the only reason people erect
>> paraphyletic groups is because they share characters that are
>> transformed in the groups detached from them. And that's a grade. So in
>> practice there is no distinction. It would certainly be possible to
>> define paraphyletic groups that weren't grades (amniota minus Carnivora,
>> for example), but who would do that?
>
> You just did. In fact to understand you statement I had to contemplate
> that group.
>
Now that I think on it, that's a grade too, defined by the absence of
> In article<c8udncWehc-...@giganews.com>,
> John Harshman<jhar...@pacbell.net> wrote:
>
>> Technically, you might be correct. But the only reason people erect
>> paraphyletic groups is because they share characters that are
>> transformed in the groups detached from them. And that's a grade. So in
>> practice there is no distinction. It would certainly be possible to
>> define paraphyletic groups that weren't grades (amniota minus Carnivora,
>> for example), but who would do that?
>
> You just did. In fact to understand you statement I had to contemplate
> that group.
>
carnassials.
pnyikos
Apr 26, 2013, 8:44:11 PM4/26/13
to nyi...@bellsouth.net
Today it's Richard's turn to get my one post of the day to this
thread. This is my second reply to this post of his, to the part I
couldn't get around to the first time.
On Apr 22, 6:01�pm, Richard Norman <r_s_nor...@comcast.net> wrote:
> On Mon, 22 Apr 2013 10:30:32 -0700 (PDT), pnyikos
methods to delineate clades, and thereby to also delineate the
paraphyletic taxa which result from pruning a clade or two or more
from the "top" of a given clade. My beef is with cladophilia, the
refusal to countenance anything except clades in formal
classifications.
> If you doubt the "violence" (yes I deliberately chose a very loaded
> term) of the cladistics revolution, look at criticisms such as
> �http://www.kheper.net/evolution/systematics/cladistics.htm
Thanks. Nice to see someone else raging against the dying of the
light, even though the site is a dozen years old. His "about me" was
updated less than half a year ago, so I'm tempted to try and contact
him when I have more free time on my hands.
He's quirky, to say the least, but that never deterred me from having
cordial relations with someone.
> and look at Cracraft's contemporary (1981) paper essentially urging
> palontologists to accept cladistics and outlining the reaction against
> it
Very one-sidedly, I'm sorry to say.
> http://research.amnh.org/vz/ornithology/pdfs/1981e.Pattern%20process%20in%20paleobiology.pdf
While I can highlight parts of the text, I cannot copy-paste what I
highlight. So my quotations will have to be brief today.
The opening paragraph already suggests that Cracraft never fathomed
Romer's thinking in _Vertebrate Paleontology_. His few references to
the book that I've seen so far do not alleviate that impression.
The following excerpt is naive at best, and suggests that Cracraft
never tried to have a respectful discussion with a systematist of the
old school:
"Several reasons can be suggested for the adverse reaction of
paleontologists towards cladistics. One is that cladists occasionally
have made provocative statements which naturally released negative
reactions"
Occasionally, my foot! Cracraft seems blissfully unaware of how
provocative his relentless repetition of the pejorative "unnatural" to
describe paraphyletic taxa is. [He even uses the word "nonexistent"
enclosed in scare quotes.] Nor does he seem to realize that he is
stacking the deck with phrases like
"genealogical groups" [read: clades]
"real units of nature" [read: clades]
"taxa lacking a unified evolutionary history" [read: paraphyletic
taxa]
The following excerpt especially left a bad taste in my mouth:
"Any classificatory procedure, such as that of evolutionary
systematics, which creates groups that are not monophyletic, will
confound analysis of these biological patterns because some of the
groups may not have a unitary genealogical history -- and it is real
history, after all, that we are trying to understand."
The part after the "because" only seems to restate what went before,
in different (and tendentious) words. What never seems to occur to
Cracraft is that the construction of phylogenetic trees could (and
did) coexist with classifications incorprorating paraphyletic taxa. A
careful look at a diagram in Romer's text near the very beginning
could have completely cured him of this blindness.
Another annoying excerpt came very soon after the two references to
Romer's text that I've seen so far. [Word searches don't work in the
website you linked.]
"If these genera were classified according to their phylogenetic
relationships, it may be that the general conclusions of such studies
would be unaffected, but perhaps not."
If he can't come up with a single example he considers to fall under
"perhaps not", he would do well to desist from such comments.
That's all I have time for until Monday, except to note that he seems
to have completely misread Gould on the topic of the excruciatingly
elementary concept of vicariance, intelligible to anyone who knows as
much about geography as I did when I was eight years old.
http://en.wikipedia.org/wiki/Vicariance
thread. This is my second reply to this post of his, to the part I
couldn't get around to the first time.
On Apr 22, 6:01�pm, Richard Norman <r_s_nor...@comcast.net> wrote:
> On Mon, 22 Apr 2013 10:30:32 -0700 (PDT), pnyikos
> <nyik...@bellsouth.net> wrote:
> >On Apr 18, 10:51�pm, Richard Norman <r_s_nor...@comcast.net> wrote:
>�That some "old" or
> >On Apr 18, 10:51�pm, Richard Norman <r_s_nor...@comcast.net> wrote:
> "traditional" �grades are monophyletic while others are polyphyletic
> really doesn't matter. �The point is whether or not the grouping is
> defined by modern phologenetic systematics ("cladistics").
What do you mean by "defined"? I have no quarrel with using cladistic
> really doesn't matter. �The point is whether or not the grouping is
> defined by modern phologenetic systematics ("cladistics").
methods to delineate clades, and thereby to also delineate the
paraphyletic taxa which result from pruning a clade or two or more
from the "top" of a given clade. My beef is with cladophilia, the
refusal to countenance anything except clades in formal
classifications.
> If you doubt the "violence" (yes I deliberately chose a very loaded
> term) of the cladistics revolution, look at criticisms such as
> �http://www.kheper.net/evolution/systematics/cladistics.htm
light, even though the site is a dozen years old. His "about me" was
updated less than half a year ago, so I'm tempted to try and contact
him when I have more free time on my hands.
He's quirky, to say the least, but that never deterred me from having
cordial relations with someone.
> and look at Cracraft's contemporary (1981) paper essentially urging
> palontologists to accept cladistics and outlining the reaction against
> it
> http://research.amnh.org/vz/ornithology/pdfs/1981e.Pattern%20process%20in%20paleobiology.pdf
While I can highlight parts of the text, I cannot copy-paste what I
highlight. So my quotations will have to be brief today.
The opening paragraph already suggests that Cracraft never fathomed
Romer's thinking in _Vertebrate Paleontology_. His few references to
the book that I've seen so far do not alleviate that impression.
The following excerpt is naive at best, and suggests that Cracraft
never tried to have a respectful discussion with a systematist of the
old school:
"Several reasons can be suggested for the adverse reaction of
paleontologists towards cladistics. One is that cladists occasionally
have made provocative statements which naturally released negative
reactions"
Occasionally, my foot! Cracraft seems blissfully unaware of how
provocative his relentless repetition of the pejorative "unnatural" to
describe paraphyletic taxa is. [He even uses the word "nonexistent"
enclosed in scare quotes.] Nor does he seem to realize that he is
stacking the deck with phrases like
"genealogical groups" [read: clades]
"real units of nature" [read: clades]
"taxa lacking a unified evolutionary history" [read: paraphyletic
taxa]
The following excerpt especially left a bad taste in my mouth:
"Any classificatory procedure, such as that of evolutionary
systematics, which creates groups that are not monophyletic, will
confound analysis of these biological patterns because some of the
groups may not have a unitary genealogical history -- and it is real
history, after all, that we are trying to understand."
The part after the "because" only seems to restate what went before,
in different (and tendentious) words. What never seems to occur to
Cracraft is that the construction of phylogenetic trees could (and
did) coexist with classifications incorprorating paraphyletic taxa. A
careful look at a diagram in Romer's text near the very beginning
could have completely cured him of this blindness.
Another annoying excerpt came very soon after the two references to
Romer's text that I've seen so far. [Word searches don't work in the
website you linked.]
"If these genera were classified according to their phylogenetic
relationships, it may be that the general conclusions of such studies
would be unaffected, but perhaps not."
If he can't come up with a single example he considers to fall under
"perhaps not", he would do well to desist from such comments.
That's all I have time for until Monday, except to note that he seems
to have completely misread Gould on the topic of the excruciatingly
elementary concept of vicariance, intelligible to anyone who knows as
much about geography as I did when I was eight years old.
http://en.wikipedia.org/wiki/Vicariance
John Harshman
Apr 26, 2013, 8:58:15 PM4/26/13
to
And I'm amazed that your only tactic is to point at what he says and
claim that it's a bad thing without giving any actual reasons. I'm
pretty sure that Joel knew that phylogenetic trees can coexist with
paraphyletic groups; his point was that such groups are not legitimate.
Just how did Joel misunderstand vicariance (a subject on which Gould
never worked and had little to say as far as I can remember)? I find it
hard to believe he was unclear on its meaning, since it's something he
has worked on.
There's no particular reason to keep stressing how clever you were at
age 8, by the way.
pnyikos
Apr 26, 2013, 9:42:59 PM4/26/13
to nyi...@bellsouth.net
I can afford to waste a few more minutes before I call it quits until
Monday.
> >>http://research.amnh.org/vz/ornithology/pdfs/1981e.Pattern%20process%...
Why aren't you amazed at the ability of atheists and others to beat on
the Bible from 20 centuries or more ago?
> And I'm amazed that your only tactic is to point at what he says and
> claim that it's a bad thing without giving any actual reasons.
I'm amazed that you can flunk reading comprehension to this extent.
Wait, it gets worse:
[snip for focus:]
> Just how did Joel misunderstand vicariance (a subject on which Gould
> never worked and had little to say as far as I can remember)?
What I *actually* wrote was that he seemed to have misread the passage
from Gould that he quoted:
"I simply cannot discern any revolutionary import in the notion of
vicariance. It is a necessary deduction from the principles and facts
of plate tectonics. The real revolution was prior and geological --
the theory of plate tectonics" (Gould 1980, p. 109)
Gould's point seems to be that the deep science behind vicariance is
plate tectonics, without which continental drift would still be widely
disbelieved by geologists [keyword: Wegener]; and hence the real cause
of most instances of vicariance would still be a mystery.
> I find it
> hard to believe he was unclear on its meaning, since it's something he
> has worked on.
Epic fail.
Click on the url and read what Cracraft wrote about Gould's passage.
It has nothing to do with either of them being unclear as to the
*meaning*.
http://research.amnh.org/vz/ornithology/pdfs/1981e.Pattern%20process%20in%20paleobiology.pdf
I don't have time to type out all the things which Joel writes on
vicariance that miss Gould's point.
> There's no particular reason to keep stressing how clever you were at
> age 8, by the way.
Actually I was stressing how crashingly elementary the concept of
vicariance is. Don't you think YOU could have understood the
*concept* [= "notion" as Gould put it] back at that age or even
earlier?
Peter Nyikos
Monday.
the Bible from 20 centuries or more ago?
> And I'm amazed that your only tactic is to point at what he says and
> claim that it's a bad thing without giving any actual reasons.
Wait, it gets worse:
[snip for focus:]
> Just how did Joel misunderstand vicariance (a subject on which Gould
> never worked and had little to say as far as I can remember)?
from Gould that he quoted:
"I simply cannot discern any revolutionary import in the notion of
vicariance. It is a necessary deduction from the principles and facts
of plate tectonics. The real revolution was prior and geological --
the theory of plate tectonics" (Gould 1980, p. 109)
Gould's point seems to be that the deep science behind vicariance is
plate tectonics, without which continental drift would still be widely
disbelieved by geologists [keyword: Wegener]; and hence the real cause
of most instances of vicariance would still be a mystery.
> I find it
> hard to believe he was unclear on its meaning, since it's something he
> has worked on.
Click on the url and read what Cracraft wrote about Gould's passage.
It has nothing to do with either of them being unclear as to the
*meaning*.
http://research.amnh.org/vz/ornithology/pdfs/1981e.Pattern%20process%20in%20paleobiology.pdf
I don't have time to type out all the things which Joel writes on
vicariance that miss Gould's point.
> There's no particular reason to keep stressing how clever you were at
> age 8, by the way.
vicariance is. Don't you think YOU could have understood the
*concept* [= "notion" as Gould put it] back at that age or even
earlier?
Peter Nyikos
John Harshman
Apr 26, 2013, 9:57:06 PM4/26/13
to
bible. But I suppose it's that the bible is still actually a current
issue, while Joel's 1981 paper isn't by any stretch.
>> And I'm amazed that your only tactic is to point at what he says and
>> claim that it's a bad thing without giving any actual reasons.
>
> I'm amazed that you can flunk reading comprehension to this extent.
>
> Wait, it gets worse:
>
> [snip for focus:]
>
>> Just how did Joel misunderstand vicariance (a subject on which Gould
>> never worked and had little to say as far as I can remember)?
>
> What I *actually* wrote was that he seemed to have misread the passage
> from Gould that he quoted:
>
> "I simply cannot discern any revolutionary import in the notion of
> vicariance. It is a necessary deduction from the principles and facts
> of plate tectonics. The real revolution was prior and geological --
> the theory of plate tectonics" (Gould 1980, p. 109)
>
> Gould's point seems to be that the deep science behind vicariance is
> plate tectonics, without which continental drift would still be widely
> disbelieved by geologists [keyword: Wegener]; and hence the real cause
> of most instances of vicariance would still be a mystery.
by the way; the only sort of vicariance that continental drift
introduces is the separation of continents by new ocean basins. Most
vicariance is much simpler: mountains, forests, plains, and such.)
>> I find it
>> hard to believe he was unclear on its meaning, since it's something he
>> has worked on.
>
> Epic fail.
>
> Click on the url and read what Cracraft wrote about Gould's passage.
> It has nothing to do with either of them being unclear as to the
> *meaning*.
>
> http://research.amnh.org/vz/ornithology/pdfs/1981e.Pattern%20process%20in%20paleobiology.pdf
>
>
> I don't have time to type out all the things which Joel writes on
> vicariance that miss Gould's point.
>> There's no particular reason to keep stressing how clever you were at
>> age 8, by the way.
>
> Actually I was stressing how crashingly elementary the concept of
> vicariance is. Don't you think YOU could have understood the
> *concept* [= "notion" as Gould put it] back at that age or even
> earlier?
Richard Norman
Apr 26, 2013, 11:23:38 PM4/26/13
to
On Fri, 26 Apr 2013 17:44:11 -0700 (PDT), pnyikos
<nyi...@bellsouth.net> wrote:
>Today it's Richard's turn to get my one post of the day to this
>thread. This is my second reply to this post of his, to the part I
>couldn't get around to the first time.
>
>On Apr 22, 6:01�pm, Richard Norman <r_s_nor...@comcast.net> wrote:
>> On Mon, 22 Apr 2013 10:30:32 -0700 (PDT), pnyikos
>> <nyik...@bellsouth.net> wrote:
>> >On Apr 18, 10:51�pm, Richard Norman <r_s_nor...@comcast.net> wrote:
>
>>�That some "old" or
>> "traditional" �grades are monophyletic while others are polyphyletic
>> really doesn't matter. �The point is whether or not the grouping is
>> defined by modern phologenetic systematics ("cladistics").
>
>What do you mean by "defined"? I have no quarrel with using cladistic
>methods to delineate clades, and thereby to also delineate the
>paraphyletic taxa which result from pruning a clade or two or more
>from the "top" of a given clade. My beef is with cladophilia, the
>refusal to countenance anything except clades in formal
>classifications.
"delineate", "define" -- is there really an important difference? No
<nyi...@bellsouth.net> wrote:
>Today it's Richard's turn to get my one post of the day to this
>thread. This is my second reply to this post of his, to the part I
>couldn't get around to the first time.
>
>On Apr 22, 6:01�pm, Richard Norman <r_s_nor...@comcast.net> wrote:
>> On Mon, 22 Apr 2013 10:30:32 -0700 (PDT), pnyikos
>> <nyik...@bellsouth.net> wrote:
>> >On Apr 18, 10:51�pm, Richard Norman <r_s_nor...@comcast.net> wrote:
>
>>�That some "old" or
>> "traditional" �grades are monophyletic while others are polyphyletic
>> really doesn't matter. �The point is whether or not the grouping is
>> defined by modern phologenetic systematics ("cladistics").
>
>What do you mean by "defined"? I have no quarrel with using cladistic
>methods to delineate clades, and thereby to also delineate the
>paraphyletic taxa which result from pruning a clade or two or more
>from the "top" of a given clade. My beef is with cladophilia, the
>refusal to countenance anything except clades in formal
>classifications.
doubt you will proceed to go on and on about how crucial this
difference is.
Your beef is noted. Too bad, you lost that argment. It is done and
over with.
>> If you doubt the "violence" (yes I deliberately chose a very loaded
>> term) of the cladistics revolution, look at criticisms such as
>> �http://www.kheper.net/evolution/systematics/cladistics.htm
>
>Thanks. Nice to see someone else raging against the dying of the
>light, even though the site is a dozen years old. His "about me" was
>updated less than half a year ago, so I'm tempted to try and contact
>him when I have more free time on my hands.
>
>He's quirky, to say the least, but that never deterred me from having
>cordial relations with someone.
>
>> and look at Cracraft's contemporary (1981) paper essentially urging
>> palontologists to accept cladistics and outlining the reaction against
>> it
>
>Very one-sidedly, I'm sorry to say.
>
>> http://research.amnh.org/vz/ornithology/pdfs/1981e.Pattern%20process%20in%20paleobiology.pdf
>
>While I can highlight parts of the text, I cannot copy-paste what I
>highlight. So my quotations will have to be brief today.
>
The point was to show that there really were strong issues at that
time. That time is done and gone and we are past all that. Well,
most of us are past that. We understand your objection to cladism. It
won't change anything or anybody. Your battles were fought with far
greater precision and force by professional working in the field and
your side lost.
The important details of phylogeny, of how evolution has proceeded in
the past, is by determining the features by which a member of some
ancestral group changed (looking at each character separately) to
found a new grouping, a new clade. At a later time, some member of
that clade developed a different change to form a new clade which is a
subset of its parent: Clade A produces subset Clade B. There is no
significance evolutionary whatsoever in considering the members of A
that are not contained in B, in set theoretic notation the set A-B.
There may be practical convenience in addressing them but there is no
reason to consider them as an "entity" evolutionarily whereas A and B
separately do have evolutionary significance.
pnyikos
Apr 29, 2013, 11:09:37 PM4/29/13
to nyi...@bellsouth.net
This thread has gotten away from talk about turtles. I'm giving it a
tiny nudge back in that direction.
On Apr 15, 9:57�pm, John Harshman <jharsh...@pacbell.net> wrote:
> On 4/15/13 9:59 AM, pnyikos wrote:
> > On Apr 13, 9:29 am, John Harshman<jharsh...@pacbell.net> �wrote:
> >> On 4/12/13 8:59 PM, pnyikos wrote:
>
> >>> On Apr 12, 10:32 pm, Richard Norman<r_s_nor...@comcast.net> � �wrote:
> >>>> On Fri, 12 Apr 2013 18:58:39 -0700 (PDT), pnyikos
> >>>> <nyik...@bellsouth.net> � �wrote:
> >>>>> On Apr 10, 6:51 pm, John Harshman<jharsh...@pacbell.net> � �wrote:
> >>>>>> On 4/10/13 3:30 PM, pnyikos wrote:
> >>>>> A number of months ago I read an article by Gauthier in which he
> >>>>> described his latest morphological cladistic analysis, in which
> >>>>> extinct as well as extant species were incuded. �Among them was the
> >>>>> enigmatic *Eunotosaurus* as well as the earliest known Chelonian. �And
> >>>>> the result was that turtles were classed as anapsids.
[big snip]
> >>> "At least two total evidence analyses (in which molecular and
> >>> morphological data are combined) suggest that turtles are not
> >>> diapsids
> >>> (Lee, 2001; Frost et al., 2006). In both, the molecular characters
> >>> are
> >>> much more numerous than the morphological ones, which implies
> >>> significant support for the placement of turtles outside diapsids in
> >>> these molecular datasets."
>
> >>> I can't see why Harshman had difficulty with this passage.
>
> >> Harshman didn't have difficulty. *You* had difficulty.
>
> > Looks to me like we both had difficulties, but of a different nature.
>
> We'll see.
>
> >>> To me it
> >>> seems perfectly straightforward: if the molecular evidence were
> >>> overwhelmingly in favor of turtles being diapsids then their sheer
> >>> number would have caused these two analyses to put turtles into
> >>> Diapsida. �But in fact they did not do that, so the molecular evidence
> >>> would seem to be only slightly in favor of them being diapsids.
>
> >> That isn't what the passage says.
>
> > It *is* what the passage literally *says*; what you talk about below
> > is what you think it *ought to* have said.
>
> Yes, I suppose so.
>
> >> There are many possible reasons why
> >> molecular evidence might fail to be decisive. The authors picked one. I
> >> mentioned others. Let's try again. First, morphological data are usually
> >> pruned of parsimony-uninformative characters, mostly invariant and
> >> autapomorphic ones, before analysis. DNA sequence data are not. So the
> >> simple size of the data set is deceptive, and the sequence data set is
> >> effectively smaller than it looks.
>
> > Huh? �When you prune characters, you reduce the number in the
> > analysis. �Didn't you mean to say "effectively bigger than it looks?"
>
> No. I meant effectively smaller, since in comparison with the
> morphological data a large proportion of the characters are useless.
Sorry, I don't understand this at all. If morphological characters
are given equal weight with nucleotide pairs, what meaning can one
attach to this or that character being useless? Even if one sequence
character cancels out another character, the "uncanceled" characters
can be expected to still be numerous enough to overshadow the
morphological ones, no?
By the way, you have a habit of sprinkling your replies to me with
sentences like the first one in the preceding paragraph, and hardly
ever trying to tell me where your difficulties lie, like I am trying
in the rest of it. If you made a little more of an effort in
explaining what you find difficult, we could probably make more
headway.
> >> Second, there is often much noise,
> >> i.e. characters some of which match all possible topologies. What the
> >> authors allege is something else: a strong, secondary signal nearly
> >> equal to the primary signal, causing conflict that renders the
> >> conclusion vulnerable to small perturbations.
>
> > Interesting. Is this somehow related to the bootstrap values for those
> > groupings being so sensitive to the inclusion of one more taxon in the
> > passage I quoted from long ago?
>
> Probably not. I don't know what caused that sensitivity. It's possible
> that the use of neighbor-joining had something to do with it.
Well, then, could you explain what "strong, secondary signals" these
authors were writing about? I mean, the authors cited by Gauthier.
Peter Nyikos
tiny nudge back in that direction.
On Apr 15, 9:57�pm, John Harshman <jharsh...@pacbell.net> wrote:
> On 4/15/13 9:59 AM, pnyikos wrote:
> > On Apr 13, 9:29 am, John Harshman<jharsh...@pacbell.net> �wrote:
> >> On 4/12/13 8:59 PM, pnyikos wrote:
>
> >>> On Apr 12, 10:32 pm, Richard Norman<r_s_nor...@comcast.net> � �wrote:
> >>>> On Fri, 12 Apr 2013 18:58:39 -0700 (PDT), pnyikos
> >>>> <nyik...@bellsouth.net> � �wrote:
> >>>>> On Apr 10, 6:51 pm, John Harshman<jharsh...@pacbell.net> � �wrote:
> >>>>>> On 4/10/13 3:30 PM, pnyikos wrote:
> >>>>> A number of months ago I read an article by Gauthier in which he
> >>>>> described his latest morphological cladistic analysis, in which
> >>>>> extinct as well as extant species were incuded. �Among them was the
> >>>>> enigmatic *Eunotosaurus* as well as the earliest known Chelonian. �And
> >>>>> the result was that turtles were classed as anapsids.
[big snip]
> >>> "At least two total evidence analyses (in which molecular and
> >>> morphological data are combined) suggest that turtles are not
> >>> diapsids
> >>> (Lee, 2001; Frost et al., 2006). In both, the molecular characters
> >>> are
> >>> much more numerous than the morphological ones, which implies
> >>> significant support for the placement of turtles outside diapsids in
> >>> these molecular datasets."
>
> >>> I can't see why Harshman had difficulty with this passage.
>
> >> Harshman didn't have difficulty. *You* had difficulty.
>
> > Looks to me like we both had difficulties, but of a different nature.
>
> We'll see.
>
> >>> To me it
> >>> seems perfectly straightforward: if the molecular evidence were
> >>> overwhelmingly in favor of turtles being diapsids then their sheer
> >>> number would have caused these two analyses to put turtles into
> >>> Diapsida. �But in fact they did not do that, so the molecular evidence
> >>> would seem to be only slightly in favor of them being diapsids.
>
> >> That isn't what the passage says.
>
> > It *is* what the passage literally *says*; what you talk about below
> > is what you think it *ought to* have said.
>
> Yes, I suppose so.
>
> >> There are many possible reasons why
> >> molecular evidence might fail to be decisive. The authors picked one. I
> >> mentioned others. Let's try again. First, morphological data are usually
> >> pruned of parsimony-uninformative characters, mostly invariant and
> >> autapomorphic ones, before analysis. DNA sequence data are not. So the
> >> simple size of the data set is deceptive, and the sequence data set is
> >> effectively smaller than it looks.
>
> > Huh? �When you prune characters, you reduce the number in the
> > analysis. �Didn't you mean to say "effectively bigger than it looks?"
>
> No. I meant effectively smaller, since in comparison with the
> morphological data a large proportion of the characters are useless.
Sorry, I don't understand this at all. If morphological characters
are given equal weight with nucleotide pairs, what meaning can one
attach to this or that character being useless? Even if one sequence
character cancels out another character, the "uncanceled" characters
can be expected to still be numerous enough to overshadow the
morphological ones, no?
By the way, you have a habit of sprinkling your replies to me with
sentences like the first one in the preceding paragraph, and hardly
ever trying to tell me where your difficulties lie, like I am trying
in the rest of it. If you made a little more of an effort in
explaining what you find difficult, we could probably make more
headway.
> >> Second, there is often much noise,
> >> i.e. characters some of which match all possible topologies. What the
> >> authors allege is something else: a strong, secondary signal nearly
> >> equal to the primary signal, causing conflict that renders the
> >> conclusion vulnerable to small perturbations.
>
> > Interesting. Is this somehow related to the bootstrap values for those
> > groupings being so sensitive to the inclusion of one more taxon in the
> > passage I quoted from long ago?
>
> Probably not. I don't know what caused that sensitivity. It's possible
> that the use of neighbor-joining had something to do with it.
Well, then, could you explain what "strong, secondary signals" these
authors were writing about? I mean, the authors cited by Gauthier.
Peter Nyikos
pnyikos
Apr 29, 2013, 11:45:55 PM4/29/13
to nyi...@bellsouth.net
On Apr 26, 11:23�pm, Richard Norman <r_s_nor...@comcast.net> wrote:
> On Fri, 26 Apr 2013 17:44:11 -0700 (PDT), pnyikos
>
>
>
>
>
> <nyik...@bellsouth.net> wrote:
> >Today it's Richard's turn to get my one post of the day to this
> >thread. �This is my second reply to this post of his, to the part I
> >couldn't get around to the first time.
>
> >On Apr 22, 6:01�pm, Richard Norman <r_s_nor...@comcast.net> wrote:
> >> On Mon, 22 Apr 2013 10:30:32 -0700 (PDT), pnyikos
> >> <nyik...@bellsouth.net> wrote:
> >> >On Apr 18, 10:51�pm, Richard Norman <r_s_nor...@comcast.net> wrote:
>
> >>�That some "old" or
> >> "traditional" �grades are monophyletic while others are polyphyletic
> >> really doesn't matter. �The point is whether or not the grouping is
> >> defined by modern phologenetic systematics ("cladistics").
>
> >What do you mean by "defined"? �I have no quarrel with using cladistic
> >methods to delineate clades, and thereby to also delineate the
> >paraphyletic taxa which result from pruning a clade or two or more
> >from the "top" of �a given clade. �My beef is with cladophilia, the
> >refusal to countenance anything except clades in formal
> >classifications.
>
> "delineate", "define" -- is there really an important difference? �No
> doubt you will proceed to go on and on about how crucial this
> difference is.
I know what I mean by "delineate". I don't know what you meant by
> On Fri, 26 Apr 2013 17:44:11 -0700 (PDT), pnyikos
>
>
>
>
>
> <nyik...@bellsouth.net> wrote:
> >Today it's Richard's turn to get my one post of the day to this
> >thread. �This is my second reply to this post of his, to the part I
> >couldn't get around to the first time.
>
> >On Apr 22, 6:01�pm, Richard Norman <r_s_nor...@comcast.net> wrote:
> >> On Mon, 22 Apr 2013 10:30:32 -0700 (PDT), pnyikos
> >> <nyik...@bellsouth.net> wrote:
> >> >On Apr 18, 10:51�pm, Richard Norman <r_s_nor...@comcast.net> wrote:
>
> >>�That some "old" or
> >> "traditional" �grades are monophyletic while others are polyphyletic
> >> really doesn't matter. �The point is whether or not the grouping is
> >> defined by modern phologenetic systematics ("cladistics").
>
> >What do you mean by "defined"? �I have no quarrel with using cladistic
> >methods to delineate clades, and thereby to also delineate the
> >paraphyletic taxa which result from pruning a clade or two or more
> >from the "top" of �a given clade. �My beef is with cladophilia, the
> >refusal to countenance anything except clades in formal
> >classifications.
>
> "delineate", "define" -- is there really an important difference? �No
> doubt you will proceed to go on and on about how crucial this
> difference is.
"define." Until I do, your "no doubt" bit will have to go begging.
> Your beef is noted. �Too bad, you lost that argment. �It is done and
> over with.
daughter's textbook in comparative vertebrate anatomy stop putting
MORE traditional taxa [most of them clades, but some of them
paraphyletic] than "modern" taxa at the ends of their books.
I forget the date on that text, but I think it is no more than five
years old.
>
>
>
>
> >> If you doubt the "violence" (yes I deliberately chose a very loaded
> >> term) of the cladistics revolution, look at criticisms such as
> >> �http://www.kheper.net/evolution/systematics/cladistics.htm
>
> >Thanks. �Nice to see someone else raging against the dying of the
> >light, �even though the site is a dozen years old. �His "about me" was
> >updated less than half a year ago, so I'm tempted to try and contact
> >him when I have more free time on my hands.
>
> >He's quirky, to say the least, but that never deterred me from having
> >cordial relations with someone.
>
> >> and look at Cracraft's contemporary (1981) paper essentially urging
> >> palontologists to accept cladistics and outlining the reaction against
> >> it
>
> >Very one-sidedly, I'm sorry to say.
>
> >>http://research.amnh.org/vz/ornithology/pdfs/1981e.Pattern%20process%...
>
>
>
> >> If you doubt the "violence" (yes I deliberately chose a very loaded
> >> term) of the cladistics revolution, look at criticisms such as
> >> �http://www.kheper.net/evolution/systematics/cladistics.htm
>
> >Thanks. �Nice to see someone else raging against the dying of the
> >light, �even though the site is a dozen years old. �His "about me" was
> >updated less than half a year ago, so I'm tempted to try and contact
> >him when I have more free time on my hands.
>
> >He's quirky, to say the least, but that never deterred me from having
> >cordial relations with someone.
>
> >> and look at Cracraft's contemporary (1981) paper essentially urging
> >> palontologists to accept cladistics and outlining the reaction against
> >> it
>
> >Very one-sidedly, I'm sorry to say.
>
>
> >While I can highlight parts of the text, I cannot copy-paste what I
> >highlight. �So my quotations will have to be brief today.
>
> <snip pnikos' criticism of Cracraft>
>
> The point was to show that there really were strong issues at that
> time. �That time is done and gone and we are past all that. �Well,
> most of us are past that. �We understand your objection to cladism. It
> won't change anything or anybody. �Your battles were fought with far
> greater precision and force by professional working in the field and
> your side lost.
Cracraft makes it seem otherwise. His opponents are strawmen.
> >While I can highlight parts of the text, I cannot copy-paste what I
> >highlight. �So my quotations will have to be brief today.
>
> <snip pnikos' criticism of Cracraft>
>
> The point was to show that there really were strong issues at that
> time. �That time is done and gone and we are past all that. �Well,
> most of us are past that. �We understand your objection to cladism. It
> won't change anything or anybody. �Your battles were fought with far
> greater precision and force by professional working in the field and
> your side lost.
As a result of the victories of your side, I'm afraid a lot of fun has
been taken out of paleontology. It's getting harder and harder to
find a university around here that teaches it. My university stopped
classes in it long ago. Clemson used to have courses in paleontology
and an undergraduate club devoted to it. Now the courses are gone,
the club has been disbanded due to lack of interest, and the only
paleontologist there now is a mere Lecturer.
He wrote to me saying there are some courses in neighboring states but
some of them are "not reputable". I haven't yet asked him what he
meant by that. I wonder if it is because they have mavericks like me
and Feduccia teaching them.
> The important details of phylogeny, of how evolution has proceeded in
> the past, is by determining the features by which a member of some
> ancestral group changed (looking at each character separately) to
> found a new grouping, a new clade. �At a later time, some member of
> that clade developed a different change to form a new clade which is a
> subset of its parent: �Clade A produces subset Clade B. �There is no
> significance evolutionary whatsoever in considering �the members of A
> that are not contained in B, in set theoretic notation the set A-B.
small subtaxon of A-B with known fossil representatives. Cladophiles
do away with those small taxons, because they are of necessity
paraphyletic. To make matters worse, they have arbitrarily decided to
reject the concept of "candidate ancestral species" and so the very
word "ancestor" might as well be labeled "unnatural" -- the pejorative
word with which they (including Harshman) label all paraphyletic
taxa.
John Harshman
Apr 29, 2013, 11:49:25 PM4/29/13
to
point yet. The reason so many molecular characters are useless is that a
great many of them are invariant; that is, the same state occurs in
every species sampled. And a great many more are autapomorphies, i.e.
the same state occurs in every species sampled but one. The same would
be true for a morphological data set sampled without bias, but nobody
codes invariant characters and few code autapomorphies. The "canceling"
you speak of clearly exists too, but it does so just about as much in
morphological characters, so there is no difference there.
> By the way, you have a habit of sprinkling your replies to me with
> sentences like the first one in the preceding paragraph, and hardly
> ever trying to tell me where your difficulties lie, like I am trying
> in the rest of it. If you made a little more of an effort in
> explaining what you find difficult, we could probably make more
> headway.
sentence would be "No.", but that doesn't seem applicable. So I don't
understand your complaint.
>>>> Second, there is often much noise,
>>>> i.e. characters some of which match all possible topologies. What the
>>>> authors allege is something else: a strong, secondary signal nearly
>>>> equal to the primary signal, causing conflict that renders the
>>>> conclusion vulnerable to small perturbations.
>>
>>> Interesting. Is this somehow related to the bootstrap values for those
>>> groupings being so sensitive to the inclusion of one more taxon in the
>>> passage I quoted from long ago?
>>
>> Probably not. I don't know what caused that sensitivity. It's possible
>> that the use of neighbor-joining had something to do with it.
>
> Well, then, could you explain what "strong, secondary signals" these
> authors were writing about? I mean, the authors cited by Gauthier.
their data. I merely point out that the justification given for assuming
a strong, secondary signal is not a good one.
pnyikos
Apr 29, 2013, 11:54:38 PM4/29/13
to nyi...@bellsouth.net
On Apr 27, 11:18�am, Mark Isaak <eci...@curioustax.onomy.net> wrote:
> On 4/26/13 8:23 PM, Richard Norman wrote:
>
> > [...] We understand your [Nyikos's] objection to cladism.
>
> I understand *that* Peter objects to cladism, but does anybody,
> including Peter, understand why?
I certainly do, and I have written about why in lots of posts. Most
of them go back to the 1990's, when sci.bio.paleontology and
sci.bio.systematics had a modest but sizable number of people who
could comprehend where I was coming from. Now there is only Harshman
left who has seen most of my objections, but he is too opinionated to
try to really understand them.
> As best I can figure, Peter has a sentimental attachment to the word
> "reptile" (and a few others) and does not want to see them go.
In short, you haven't a clue as to where I am coming from.
The big question is, are you interested enough to let yourself be
informed?
>�No
> problem. �Cladism applies to formal systematics; it does not revise the
> English language. �"Reptile" will continue in use, as will "apterous
> insects", "tree", "plankton", and many other useful non-cladistic
> classifications. �Chill, Peter!
That two-word sentence suggests that the answer to my big question is
a resounding NO.
Peter Nyikos
> On 4/26/13 8:23 PM, Richard Norman wrote:
>
> > [...] We understand your [Nyikos's] objection to cladism.
>
> I understand *that* Peter objects to cladism, but does anybody,
> including Peter, understand why?
I certainly do, and I have written about why in lots of posts. Most
of them go back to the 1990's, when sci.bio.paleontology and
sci.bio.systematics had a modest but sizable number of people who
could comprehend where I was coming from. Now there is only Harshman
left who has seen most of my objections, but he is too opinionated to
try to really understand them.
> As best I can figure, Peter has a sentimental attachment to the word
> "reptile" (and a few others) and does not want to see them go.
In short, you haven't a clue as to where I am coming from.
The big question is, are you interested enough to let yourself be
informed?
>�No
> problem. �Cladism applies to formal systematics; it does not revise the
> English language. �"Reptile" will continue in use, as will "apterous
> insects", "tree", "plankton", and many other useful non-cladistic
> classifications. �Chill, Peter!
That two-word sentence suggests that the answer to my big question is
a resounding NO.
Peter Nyikos
John Harshman
Apr 30, 2013, 12:02:03 AM4/30/13
to
connection between two things you don't like from your nether regions.
You have no justification for any of this.
> It's getting harder and harder to
> find a university around here that teaches it. My university stopped
> classes in it long ago. Clemson used to have courses in paleontology
> and an undergraduate club devoted to it. Now the courses are gone,
> the club has been disbanded due to lack of interest, and the only
> paleontologist there now is a mere Lecturer.
>
> He wrote to me saying there are some courses in neighboring states but
> some of them are "not reputable". I haven't yet asked him what he
> meant by that. I wonder if it is because they have mavericks like me
> and Feduccia teaching them.
and well at a great many universities. I took three quarters worth at U.
Chicago, all fine cladistic stuff not to your liking, but all us
students were quite enthused.
>> The important details of phylogeny, of how evolution has proceeded in
>> the past, is by determining the features by which a member of some
>> ancestral group changed (looking at each character separately) to
>> found a new grouping, a new clade. At a later time, some member of
>> that clade developed a different change to form a new clade which is a
>> subset of its parent: Clade A produces subset Clade B. There is no
>> significance evolutionary whatsoever in considering the members of A
>> that are not contained in B, in set theoretic notation the set A-B.
>
> It is if we can narrow down the ancestry of the members of B to some
> small subtaxon of A-B with known fossil representatives. Cladophiles
> do away with those small taxons, because they are of necessity
> paraphyletic.
> To make matters worse, they have arbitrarily decided to
> reject the concept of "candidate ancestral species" and so the very
> word "ancestor" might as well be labeled "unnatural" -- the pejorative
> word with which they (including Harshman) label all paraphyletic
> taxa.
have no way to tell if it's really an ancestor and 2) you can't
recognize biological species in the fossil record anyway. The only
purpose served is to give you a warm, fuzzy feeling. It isn't science.
pnyikos
Apr 30, 2013, 12:03:08 AM4/30/13
to nyi...@bellsouth.net
John Harshman
Apr 30, 2013, 12:03:57 AM4/30/13
to
On 4/29/13 8:54 PM, pnyikos wrote:
> On Apr 27, 11:18 am, Mark Isaak<eci...@curioustax.onomy.net> wrote:
>> On 4/26/13 8:23 PM, Richard Norman wrote:
>>
>>> [...] We understand your [Nyikos's] objection to cladism.
>>
>> I understand *that* Peter objects to cladism, but does anybody,
>> including Peter, understand why?
>
> I certainly do, and I have written about why in lots of posts. Most
> of them go back to the 1990's, when sci.bio.paleontology and
> sci.bio.systematics had a modest but sizable number of people who
> could comprehend where I was coming from. Now there is only Harshman
> left who has seen most of my objections, but he is too opinionated to
> try to really understand them.
Translation: he thinks Peter's objections are wrong, and therefore can't
> On Apr 27, 11:18 am, Mark Isaak<eci...@curioustax.onomy.net> wrote:
>> On 4/26/13 8:23 PM, Richard Norman wrote:
>>
>>> [...] We understand your [Nyikos's] objection to cladism.
>>
>> I understand *that* Peter objects to cladism, but does anybody,
>> including Peter, understand why?
>
> I certainly do, and I have written about why in lots of posts. Most
> of them go back to the 1990's, when sci.bio.paleontology and
> sci.bio.systematics had a modest but sizable number of people who
> could comprehend where I was coming from. Now there is only Harshman
> left who has seen most of my objections, but he is too opinionated to
> try to really understand them.
possibly understand them, because they're right, dang it.
>> As best I can figure, Peter has a sentimental attachment to the word
>> "reptile" (and a few others) and does not want to see them go.
>
> In short, you haven't a clue as to where I am coming from.
tradition. All else is excuse.
pnyikos
Apr 30, 2013, 12:25:57 AM4/30/13
to nyi...@bellsouth.net
On Apr 30, 12:03�am, John Harshman <jharsh...@pacbell.net> wrote:
> On 4/29/13 8:54 PM, pnyikos wrote:
>
> > On Apr 27, 11:18 am, Mark Isaak<eci...@curioustax.onomy.net> �wrote:
> >> On 4/26/13 8:23 PM, Richard Norman wrote:
>
> >>> [...] We understand your [Nyikos's] objection to cladism.
>
> >> I understand *that* Peter objects to cladism, but does anybody,
> >> including Peter, understand why?
>
> > I certainly do, and I have written about why in lots of posts. �Most
> > of them go back to the 1990's, when sci.bio.paleontology and
> > sci.bio.systematics had a modest but sizable number of people who
> > could comprehend where I was coming from. �Now there is only Harshman
> > left who has seen most of my objections, but he is too opinionated to
> > try to really understand them.
>
> Translation: he thinks Peter's objections are wrong,
...but is unable to explain why, his attempts largely consisting of
> On 4/29/13 8:54 PM, pnyikos wrote:
>
> > On Apr 27, 11:18 am, Mark Isaak<eci...@curioustax.onomy.net> �wrote:
> >> On 4/26/13 8:23 PM, Richard Norman wrote:
>
> >>> [...] We understand your [Nyikos's] objection to cladism.
>
> >> I understand *that* Peter objects to cladism, but does anybody,
> >> including Peter, understand why?
>
> > I certainly do, and I have written about why in lots of posts. �Most
> > of them go back to the 1990's, when sci.bio.paleontology and
> > sci.bio.systematics had a modest but sizable number of people who
> > could comprehend where I was coming from. �Now there is only Harshman
> > left who has seen most of my objections, but he is too opinionated to
> > try to really understand them.
>
> Translation: he thinks Peter's objections are wrong,
jeers like the use of pejorative words.
And the sad part is, he no longer realizes they are pejorative. This
is obvious from his clueless reaction to my criticism of Cracraft:
"And I'm amazed that your only tactic is to point at what he says
and claim that it's a bad thing without giving any actual reasons."
Here is how that "only tactic" actually read:
____________excerpt_______________________
The following excerpt is naive at best, and suggests that Cracraft
never tried to have a respectful discussion with a systematist of the
old school:
"Several reasons can be suggested for the adverse reaction of
paleontologists towards cladistics. One is that cladists
occasionally
have made provocative statements which naturally released negative
reactions"
Occasionally, my foot! Cracraft seems blissfully unaware of how
provocative his relentless repetition of the pejorative "unnatural"
to
describe paraphyletic taxa is. [He even uses the word "nonexistent"
enclosed in scare quotes.] Nor does he seem to realize that he is
stacking the deck with phrases like
"genealogical groups" [read: clades]
"real units of nature" [read: clades]
"taxa lacking a unified evolutionary history" [read: paraphyletic
taxa]
> >> As best I can figure, Peter has a sentimental attachment to the word
> >> "reptile" (and a few others) and does not want to see them go.
>
> > In short, you haven't a clue as to where I am coming from.
>
> I think that's what it boils down to: a sentimental attachment to
> tradition. All else is excuse.
It is loading more messages.
0 new messages