[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: Lü and [J]i 2006, a review

I've copied my apologies to the top to avoid useless hard feelings:

Sorry for my direct tone. I just want to be really sure that my point gets across. I also apologize for the likely culture shock (Americans seem to get encouragement and praise throughout their lives, while Europeans generally think that passionate people -- like you -- don't need any additional encouragement and feel that praise without a little "but" here and there cannot possibly be serious*) the impact of which I have certainly exaggerated here.

* As educational approaches, both have interesting advantages and disadvantages.

<1. Well, the choice of the outgroup taxa dooms the base of this study from the start. Pterosaurs are lizards, not archosaurs as shown
by a cladistic analysis 2 to 14 times larger and more inclusive than any
prior attempt at classifying amniotes.>

I doubt whether the outgroup would have influenced the deeper
arrangements of pterosaurs much at all, though it might shift the base around
a little. 

So, Jaime, why make a red herring negatory comment if you are in
tacit agreement with my statement? It makes you look bad, my friend.
Argumentative, just for the sake of being so.

Why? Most scientists actively look for nits to pick when given the opportunity. Attention to detail is, generally speaking, very useful in science.

  It's a "highly derived dead end" in EVERY cladistic analysis to
include it, which comes from the nature of it's peculiar cranial anatomy.


Not so here, my friend. Show me one study that has Rhamphorhynchus as
a dead end without sister progeny. It's always the last bus before
the transfer.


Basics of cladistics:

1. Cladistics is completely incapable of recognizing ancestors as such. It RECONSTRUCTS HYPOTHETICAL ancestors -- they sit at the nodes, not at the end of any terminal branch.
2. If an OTU has a zero-length branch (means, no autapomorphies) and its geological age fits, then it ENTERS CONSIDERATION in the search for an ancestor. This is, however, not proof -- we may have overlooked any number of autapomorphies that, for example, didn't fossilize.
3. From 2., it follows that science can show that something is NOT the ancestor of anything known, but it cannot show that something IS the ancestor of anything known. Science can disprove, but not prove. Therefore cladistics is completely incapable of recognizing ancestors as such.
4. Worse yet. All cladistics programs I'm aware of don't display a zero-length branch as having zero length unless you have told them to display all branch lengths. In other words, the programs are incapable of putting an OTU at a node. Every OTU, even if it truly is an ancestor of another OTU (or several), will INEVITABLY get its own terminal branch. I can also express this the other way around: if read literally, no cladogram you have ever seen contains an ancestor. The claim that there are cladograms which show *Rhamphorhynchus* as an ancestor of anything _is wrong_, because it's a contradiction in terms.

Sorry, but... I can't escape the conclusion that either I have majorly misunderstood you (in which case I apologize) or that you don't know how to read a cladogram. _In the latter case_, you are in grave danger of doing pseudoscience because _you're doing cladistics without understanding what you're doing_.

Again, sorry for my direct tone. I just want to be really sure that my point gets across. I also apologize for the likely culture shock (Americans seem to get encouragement and praise throughout their lives, while Europeans generally think that passionate people -- like you -- don't need any additional encouragement and feel that praise without a little "but" here and there cannot possibly be serious*) the impact of which I have certainly exaggerated here.

* As educational approaches, both have interesting advantages and disadvantages.

Less confusing, I
am sure, is the more plesiomorphic postcrania.


"I am sure" is a well-known weasel phrase for "I have room for for
doubt but don't want to express it."

I interpret it as "I can't read your mind, so I don't know what exactly you would find more confusing."

And the postcrania is not as plesiomorphic as you might assume. In my
study, there isn't just one Rhamphorhynchus, but a whole bush leading
from Campylognathoides alone, with lots of variation in every part of
the body.

With more or less juvenile proportions and, importantly, more or less juvenile bone histology (number of LAGs, proportion of fibrolamellar and Haversian bone, presence of unfinished bone surface, presence of external fundamental system).

It seems to have escaped you that actual serious research on pterosaur ontogeny has been done. That's odd, because I know you read the Journal of Vertebrate Paleontology. What am I missing?

so I doubt Rhamphorhynchidae is very incoherent in the current
pterosaur environment.

??? You've phrased that incoherently.

Probably in-co-herent has its literal meaning here, "not sticking together", and thus means "polyphyletic"...

  These specimens are not included for what, it seems to me, is a
VERY good reason: they are considered juveniles by just about
every pterosaur worker out there.

That's tradition and paradigm. Test them. No one has yet. 

Flat-out wrong.

Not because they assume it to be true, but because a good deal of work
has gone into considering them and relegating them as juveniles,
even if taxonomic assignment can be shaky (which it may be).

Not so, small size was assumed to represent juvenile status and long- winded explanations were devised to promote the idea despite changes in bone, tooth and proportion that could just as easily, and it turns out, more readily, to have been phylogenetic. 

See above.

If I were wrong, I should be the one with 34,000 MPTs. Right? Right? 

No, not necessarily. The number of MPTs depends on the exact data matrix.

No I have one tree,
and all the sister taxa blend, like Mr. Darwin said they would.


Like Mr. Darwin said they would _if the fossil record were complete_.

Don't rely too much on "blending", on superficial morphological intermediates. I remember well the precladistic attempts to derive Tyrannosauridae from *Allosaurus*, with e. g. *Acrocanthosaurus* as an intermediate that blended in very, very nicely. Turns out people had looked at far too few characters, even before things like *Stokesosaurus* were widely acknowledged as tyrannosauroids.

Many discoveries, from continental drift to a heliocentric solar
system, were made by mad men working alone with various sorts
pointing fingers and calling names.


Erm...

_Now_ you _really_ sound like a pseudoscientist. You know, "in order to be the next Galileo, it is not enough to be persecuted by an orthodoxy. You also have to be right" -- Carl Sagan, or wait, I forgot who said it first.

And you, Jaime, are one of the
sticks in the mud, as you'll discover someday.


I'll try to test that...

Finding proof these aren't juveniles would be considerable, but
so far, no data demonstrating this has been published.


Not for lack of trying. The strategy now is to publish and finished
revision of the Amniota, then work to the branches.


Fine -- I'll wait for the paper.

So, there is good reason
not to include these specimens in a matrix.


Yes, put your head in the sand and avoid testing when it would be so
easy to do so. And so rewarding. I encourage you to put any tiny
ptero in a matrix and see where it falls. A dozen would reveal more
though.

I encourage you to put any juvenile of anything into the same matrix as its parents. In many cases the babies will be closer to the root of the tree than their parents -- because they have not yet developed all apomorphies that their parents show. Apes would be a funny test case, with us being so paedomorphic...

Wait! One word: Aublysodontinae.

There's also a very interesting paper called "Ontogeny discombobulates phylogeny" about this problem in salamander phylogeny (Systematic Biology, February last year), but of course salamanders with all their metamorphosis are a more extreme case than any amniote.

Removing the tiny pteros does indeed bounce the cladogram back to a
ladder -- but then you have to deal with sister taxa that do not
blend as Darwin said they should.

If -- the -- fossil -- record -- were -- complete. The pterosaur fossil record is quite pathetic, as I'm certain you agree.

There is something deliciously wonderful about lifting the veil of
ignorance and seeing unexpected truth after years of work.


Stop using the word "truth", _please_.

There is also something devilishly charming about certain folk who
play their part so well and create unnecessary drama about such trivia.

Don't forget to play your own part. Go ahead and show that your tiny pterosaurs are adult. Tricking oneself into seeing doubly impossible* "ephemeral juveniles" around them does not count.

* A completely unossified terrestrial vertebrate of that size would be hyper-mega-rhachitic at best, and uncalcified cartilage cannot preserve in ordinary Konservatlagerstätten -- you'd need an oil sand at the very least, if not outright ice.