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

Re: Cladistics and stuff



As usual, Ralph shows tremendous insight and wisdom and an admirable lack of 
fear of being highly technical (e.g., pain-in-the-buttedness) when necessary to 
get his point across, even in a dinosaur discussion, where the mere thought of 
doing anything remotely quantitative tends to make people start bleeding 
internally (for the record, doing cladistics using characters like "brevis 
shelf longer than wide" isn't really quantitative with respect to the data).  
In time, we really need to move beyond discussing features in terms of them 
being "robust" or "large" if we ever hope to have systematics really deal well 
with morphological variation, which is probably going to remain the key to 
paleontological systematics and taxonomy for some time to come.  We can talk at 
great length about how to massage and interpret the 0s and 1s in our character 
matrices, but sooner or later we are
going to have to sit down and really address the variation, distribution, and 
description of the characters those digits represent.  I think it is starting 
and many people are moving toward it, but we are still quite a ways from really 
understanding character variation.  Just some random neural firings motivated 
by swimming in theropod dental characters and watching almost all of them 
collapse when their variation and distribution is assessed....

-Josh

Ralph Chapman wrote:

> Just thoughts along this line.
>
> I hate the idea that the dino/bird "controversy" sometimes gets really caught 
> up in rather detailed discussions on exact cladistic, or perhaps better 
> stated, phylogenetic, procedures, which have been used - and I'm NOT talking 
> about HP Kinman here - by some as part of a long term campaign to obfuscate 
> the real dialog about dinosaur bird origins. The weaknesses, and the 
> pain-in-the-buttedness (technical term, for those not directly in the 
> profession) of some of current phylogenetic procedures are well known to many 
> of us who are racking our brains to come up with a better system, which would 
> include HP Kinman, of course. I'm still waiting but am still thinking.
>
> In the dino/bird dialog I would just settle for phylogenetic approaches that 
> are just quantitative, repeatable and explicitly stated as a good starting 
> point for a solid dialog. Then we can argue the fine points of which 
> algorithms do what, as well as look for similarities in the results of the 
> various approaches - which suggests robustness in those results - and, of 
> course, differences, which is where the neat areas for developing discussion 
> exist. Let's indeed experiment with all sorts of use, modification and 
> perhaps even abuse of the parsimony and parsimoniesque - another technical 
> term 8-)  - assumptions of cladistic approaches and see what the heck happens.
>
> ¯---quick aside ¯- my lovely bride Linda (Deck) and some friends were in one 
> of her friends kitchen while she was in high school. They were trying to 
> figure out how to cook a frozen pizza (which should be illegal in Buffalo, NY 
> which has great real pizza). They were arguing how to do it just the right 
> way, when the grandmother of one of the friends heard the ruckus and came 
> into the kitchen and said "Just put the damn pizza in the damn oven and see 
> what the hell happens!" I see great wisdom from age there. Let's just play 
> with lots and lots of assumptions and see how robust things really are. See 
> what the problems really are.
>
> I just get tired of the anti-dino origin mafia giving nothing but classic, 
> non-quantified or even repeatable observations in what they see as the 
> phylogenetic relationships. "Oh, I have a warm ookie feeling this thing is a 
> bird and it is because I say so." Sorry, this does not cut it in science, or 
> to put it in feducciary terms, "Hypothesis non-existent!" so rejected.
>
> So, let's experiment with all sorts of different algorithms and, even more 
> important, let's actually pay attention and discuss the characters in detail! 
> Even though I do phylogenetic stuff using classic cladistic algorithms I 
> still get really sick and tired of characters that go like -
>
> orbit not round but also not sub-ovate
>
> real quality morphometric observation there (whatever face thing means 
> sarcastic should be inserted here). I really think lots of cladists reject 
> the use of detailed morphometrics because they really know that many 
> (hopefully not most) of their characters will really end up being continuous 
> blobs if actually studied with some morphometric intelligence. I think the 
> great part of Hutch's recent 2 papers (great papers in my first read) is that 
> he really concentrated on the functional/developmental implications on 
> characters, what really were different characters. We should be doing that 
> with morphometrics as well - although Miriam Zelditch and Bill Fink have been 
> leading the way here and should be congratulated whether you like their 
> specific morphometric techniques or not.
>
> So, if someone is not willing to be explicit and quantitative and repeatable 
> they really have nothing to say in this dialog except just another intuitive 
> opinion. If someone is not willing to expose their characters to real 
> heavy-duty analysis they are in a similar boat.
>
> RANT - The next time I hear someone give a single ratio value as the break in 
> a character distribution, I think I'll scream. Almost all bivariate 
> distributions have ratios that change throughout ontogenies (unless they go 
> through the origin and are linear, which is very rare), and many taxa share 
> the same ratio at some time in their growth. The power is in actually looking 
> at the bivariate relationships, the occupation of the morphospace - not the 
> ratio. -
> END 'O RANT
>
> Anyway, let's take this discussion away from the warm ookie belly feelings 
> and just get it to real quantitative discussions. Frankly, I don't care if 
> someone wants to use Dinosauria as excluding Aves as long as he/she is doing 
> a thought-out phylogenetic analysis of some sort and they are explicit about 
> what they mean. If an anti person has an analysis of coded characters - 
> explicitly stated - that are phylogenetically reconstructed using some 
> algorithm - and this is a broad term - that I can sit and do myself at my 
> desk, then we can actually have a scientific discussion. the rest is just a 
> discussion of the qualtities of the data and algorithms and that can be 
> progressive and constructive. Sure beats wearing a girdle around both your 
> shins and waddling around and thinking something scientific is actually 
> happening regarding the bird/dino debate (semi-cryptic reference here).
>
> And if the hair on pterosaurs (at least the one) is the same as dino fuzz, I 
> would be just delighted because it would really be fun and be a very 
> interesting and mischievious thing. Would not expel the dino/bird origin 
> theory, would just drop dino fuzz down a bunch of nodes.
>
> Anyway, remember the bird-dino origin stuff seems to be pretty algorithm 
> robust and it will take considerable effort to break it. Everyone should try 
> and do that, Anti-dino origin types included. the latter just need to do 
> their job much better than they have. They seem to see negative feelings from 
> most dino types as being the result of trying to disprove the dino origin 
> idea. It is not, from me at least it is just disappointment in the quality of 
> their arguments, which seems to be pretty low up to now.
>
> Triceratops exhibit almost done (2 weeks left) and I'm going insane, which is 
> probably apparent from this post. Bruce Mohn did a construction of the old 
> mount using the miniature bones we generated that is to die for! Bob Walters' 
> art is wonderful. We also have some bronzes of the miniature head that also 
> make your heart go pity-pater as well. Bought a tux and am looking forward to 
> creative black tie and removing a larger than big ungulate sized monkey off 
> all our backs.
>
> Ralph Chapman
>
> Ralph E. Chapman
> Applied Morphometrics Laboratory
> National Museum of Natural history
> ADP, EG-15  NHB, 10th & Constitution, NW
> Smithsonian Institution
> Washington, DC 20560-0136
> (202) 786-2293, Fax: (202) 357-4122
> Chapman.Ralph@nmnh.si.edu




--
Josh Smith
Department of Earth and Environmental Science
University of Pennsylvania
471 Hayden Hall
240 South 33rd Street
Philadelphia, PA  19104-6316
(215) 898-5630 (Office)
(215) 898-0964 (FAX)