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

reply to D. Naish: distrust makes good science!

Dear Mike, Darren and dinolisters:

Distrust makes good science. It's good that you distrust my observations. Make 
your own observations then apply them. See what you get.

Whether I made no mistakes, three mistakes, a dozen mistakes or a hundred 
mistakes on character coding, all YOU have to do is plug those TAXA into any 
pterosaur inhabitated cladogram and let the results speak for themselves. Use 
any characters you want to! I'm just the messenger. The taxa are the message. 

Disregard everything I've ever published. But don't disregard these candidates 
on account of my supposed mishandling them. I recall I also P.Od Ellenberger 
for suggesting that his precious bird-ancestor candidate, Cosesaurus, was a 
more probable pterosaur-ancestor candidate. So you're in good company. I wake 
up every day knowing I'm universally disliked for making new suggestions.

On the other hand, the very first Photoshop interpretation I ever attempted was 
tracing the skull of the first langobardisaur hidden beneath its ribs. The 
result was much closer to that of the second big-eyed langobardisaur than that 
which was originally figured. So some tracings do come out right.

I apologize to all you pterosaurs, fenestrasaurs, macrocnemians and lizards I 
may have misrepresented. But you little fragile guys may be partly to blame for 
often being crushed, scattered, covered in your own soft membranes and poorly 
documented. Plus, typically you're ignored by all but a few, so confirmation or 
condemnation rarely--if ever--comes.

I eagerly support any and all independent testing of any candidates I have 
proposed and still stand by them. Whenever independent testing of fenestrasaurs 
happens, someone please let me know. So far, I haven't seen it. If you guys are 
talking about my results among yourselves without including me, that's keeping 
me in the dark and out of the fraternity. How is that helping science? If you 
want to shut me up, show up with some evidence. 

If you're convinced that archosaurs with their atrophied lateral digits and 
deep chevrons make better pterosaur cousin candidates than Huehuecuetzpalli > 
Longisquama then let's see your cards. Really. And now is a good time. If you 
have in your hands evidence for a deep pterosaur wing, send it over as a jpeg, 
too. Let's talk about this. 

re: Photoshop. It is the only verifiable tool available when you need to study 
hundreds of specimens and you don't have a 'golden ticket' to go around the 
world to see them in person. It provides for hyper-accurate tracing, sharing of 
electronic files, and feedback on multiple file layers -- in other words: 
discussion with the tracer on an accurate rendition base of an image of the 
fossil itself. What could be better? And by 'verifiable' I mean: I WANT you to 
find errors so that rectifications can be made. 

If we don't use this tool, we're stuck with this: The typical professorial 
description of a fossil is a series of words strung together in paragraph form 
that indicates, as well as can be told, what one has observed. If unsupported 
by photographic documentation, the evidence must be taken on faith, trust and a 
rough outline drawing rather than 300 dpi halftone dots and a tight tracing. 
I'm only asking for higher standards in this digital age. The paragraph format 
does work with bone you can hold. Maybe it doesn't work so well when tiny bones 
and soft tissue every so often dip back into the matrix. 

DN wrote: "Every time I (or anyone else I've heard from) looks at 
pterosaurs,_Longisquama_, _Sharovipteryx_, _Huehuecuetzpalli_ or whatever, we 
come away rejecting your multiple unique interpretations."

Are you saying (like Chris Bennett has said on his web site), that because I 
supposedly missed seeing two bone edges, ALL my interpretations are suspect? 
And therefore are rejected? I remind you that in that same website Chris 
interprets a maxilla as a premaxilla without indicating where the typically 
much bigger maxilla might be. 

Or are you saying that ALL my observations are mistakes, including the big 
highly visible bones? 

I hope you're NOT saying, that because I've bungled a few observations, you 
reject the possibility that these taxa are unworthy candidates for sharing 
pterosaur DNA? (David Hone, please take note before publishing the next chapter 
of your dissertation.)

I also hope you're NOT saying that if I agree with and use the observations of 
others that they share my malady.

Trash me. Don't trash the taxa. 

I admit, as I have admitted for some time, that what I am searching for can be 
so ephemeral that illusions can appear. Resolution is sometimes not good. I do 
push for the most out of my data. I have made mistakes. But I've also 
documented more fossils in this clade than any other worker. I also wonder if 
most of my observations are illusions why some illusions are the same in sister 
taxa and why they don't jump out in cladograms as awkward exceptions and 
grevious errors?

As I've often said, in my cladograms you can delete up to a third of the 
characters at random and still arrive at the same single tree phylogeny. You 
can add taxa at will and do the same. No current phylogenetic analysis can make 
that claim. It's chronological. It creates morphological blends. That's what 
we're looking for, right?

Perhaps because my data matrix is so large and that to inspect it would be such 
a daunting challenge that the best course to avoid the work would be by 
chipping away at a few characters? I hope not.

Like any good scientist, I am simply awaiting validation or condemnation from 
an independent source who has also tested the same taxon list. Use your own 
characters! Distrust makes science.

If, as David Hone and Chris Bennett have done, a few characters are attacked in 
order to undermine the entire description and thus reject the taxon as 
unworthy, then all they have done is to attack the messenger.

Let's get back to arguing and testing in good fellowship.

David Peters
St. Louis

Forwarded from Darren, by request.

------- start of forwarded message -------
Date: Fri, 8 Jun 2007 13:00:12 +0100
From: "Darren Naish" <eotyrannus@gmail.com>
To: "Mike Taylor" <mike@indexdata.com>
Subject: On idiosyncrasy in pterosaur-land

Mike - any chance you could please fwd this to DML for me? Thanks..

WRT Dave Peters' message 'idiosyncrasy in birdland', I thought that - given
that I was charged as one of the workers ignoring 'the elephant in the room'
- I'd respond.

David: obviously your efforts to document pterosaur phylogeny and origins by
way of comprehensive character analysis are worthwhile pursuits. While there
is still some completed work to appear on how your proposed phylogenetic
schemes match up with the evidence (recall that Dave Hone referred to
unpublished material in his phd thesis), I would argue that there is a very
good reason why your phylogenies are not being accepted as the most
parsimonious: this is because your choice of characters, and coding of them,
rests on your photo-interpretation technique. I've said it before and I'll
say it again (and everyone else I've spoken to says the same) - none of us
hold any confidence in your many unique interpretations of pterosaur bones
and teeth, soft tissue crests, frills and tail-whips, multiple babies etc.
Every time I (or anyone else I've heard from) looks at pterosaurs,
_Longisquama_, _Sharovipteryx_, _Huehuecuetzpalli_ or whatever, we come away
rejecting your multiple unique interpretations. As discussed in personal
correspondence, I have found that your photo-interpretation method has led
you astray on _Istiodactylus_, _Tupuxuara_, and Solnhofen _Pterodactylus_
and _Rhamphorhynchus_ specimens, and I have heard likewise from workers with
first-hand experience of the Triassic taxa that are integral to your
prolacertiform/squamate origins hypothesis. You are coding characters and
character states that are not present.

The conclusion is that your data sets are cluttered with a significant
amount of irrelevant noise (I'm not sure how much, but recall it looking
like more than 50%), hence the lack of acceptance. Sorry, but my ideas on
this haven't changed, nor have those who have criticised this aspect of your
work in the past. And this isn't me being nasty, as I was initially _really_
impressed with your 'four prolacertiforms' paper.

All the best.

Back to storm prediction in butterflies...

Darren Naish
School of Earth & Environmental Sciences
Burnaby Building, Burnaby Rd
University of Portsmouth
Portsmouth, UK, PO1 3QL

tel: 023 92846045
visit my blog: http://scienceblogs.com/tetrapodzoology/
------- end of forwarded message -------