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

Was: Wukongopterus and Darwinopterus, now phylogenetics etc.



On Nov 29, 2009, at 6:17 AM, David Marjanovic wrote:

>> Every right tree we hypothesize will admit
>> new taxa as they are discovered with no change to the rest of the
>> tree topology.
> 
> No, why? I've seen back-and-forth changes happen with corrections, why 
> shouldn't they also happen with addition of taxa? I can't think of a good 
> example right now, but I think I've seen that happen, too...

For the moment you have complete knowledge of every living thing. You snap your 
fingers and POW the tree of life is assembled. That's the "right tree." Oops, 
there's a new taxon that just came to light, let's see it goes right here... 
voila. No change to the tree. In fact, that new taxon fills a gap left over 
from earlier. That's what I'm talking about. 


> 
>> Getting closer and closer to the The Truth should be the push.
> 
> That's not how science works.
> 
> First of all, suppose you discover the truth. How can you figure out that 
> what you've found is in fact the truth? By comparing it to the truth, which 
> you don't have?
> 
> Science is not a quest for truth, it's a quest for falsehood. It tries to 
> eliminate all possibilities that are impossible, and then all those that are 
> too improbable (unparsimonious); but there are just too many of those 
> (including lots that nobody has ever thought of) for this approach by 
> elimination to lead to a single answer with full certainty.

Funny. I wish I could tell you all the people who did NOT first spy the moons 
of Jupiter, but what does that matter? I can only tell you of Gallileo G. I'd 
like Nature and Science to publish all by flubs and flops, but they're only 
interested in new discoveries. I think you would be very entertaining company. 
We should room together at some upcoming convention. 
> 
>> You're never going to figure out the perimeter of England because you
>> come up against fractals.
> 
> But you can come close enough, either by mathematical tricks like 
> integration, or (in theory anyway) by bumping against the fact that space is 
> quantified (keyword: Planck's length).

Close enough is good enough for me. 
> 
>> This complexity you're suggesting is
>> another facet of fractals. Avoid it or delve into madness.
> 
> You can't avoid it if you don't want to get a hopelessly wrong answer. It's 
> like thermodynamics: you can't win, you can't break even, and you can't quit 
> the game. :-)

Excellent. That's why I have studiously avoided philosophy and quantum 
mechanics. Let someone else invent warp drive.


> 
>> All we're looking for are simple models for complex processes
>> taking millions of years and just as many generations.
> 
> "Things should be made as simple as possible, but not any simpler."
> -- Albert Einstein
> 
> You advocate making them simpler than possible. Inevitably you get wrong 
> answers. :-|


Again with the unfounded insults! David, if you're going to tell me I get wrong 
answers, you have to point out which are wrong (and by inference, the rest will 
be right!)  Simpler than possible? Isn't that, by definition, "impossible"?
> 
>> > Presuming you mean everything should be unordered and ignored for
>> > correlations, there are quite a few manuscripts in the cladistics
>> > literature that suggest otherwise.
>> 
>> Posting your question back to you: how did they know they were more
>> right than wrong when they concluded?
> 
> You should read more Systematic Biology and journals like that. Simulation 
> studies to see if a method works are normal these days in biology. There are 
> several papers in the June 2003 issue of JVP that explore the "problem" of 
> missing data in phylogenetic analysis by simulation.

I'm trying to avoid the abstract and concentrate on reality. 
> 
>> > However, low resolution does not mean a tree is "bad".  Sometimes,
>> > low resolution is the correct answer given the data at hand: it
>> > simply means that taxa cannot be differentiated in a repeatable
>> > manner according to the present data.  That may be the error-free,
>> > robust answer for a given dataset.
>> 
>> Then get more data. It's always there.
> 
> Except, if you don't publish for too many years, you'll get into trouble if 
> you have or want an academic job. Better publish intermediate results every 
> once in a while -- that's also a good opportunity to let others go over your 
> matrix and point out mistakes to you.

I see. Do they do that? And, going back to an earlier thread, if someone finds 
three mistakes is that enough to "distrust" a matrix with 40,000 cells? That's 
reality, my friend. 

> 
>> > Please name any apomorphy that is not rare.
> 
> Ovoviviparity and viviparity in squamates...

"Viviparity has evolved almost 100 times among lizards and snakes." : Shine and 
Gillette 1988.
> 
> Color. On scales like tetrapod phylogeny, integument color has no 
> phylogenetic signal whatsoever.

Sounds like we're back to fractals. How to grade, judge and score color, 
patterns, etcs. 

> 
> Body size. _Has_ a strong phylogenetic signal, but not as strong as it could 
> be.

Every tetrapod has "body size."  I'm making fun, but your line of reasoning is 
interesting. Hone and Benton 2007 Anurognathus and Quetzalocoatlus on a 
diagram, drew a straight line between them and came up with a Cope's Rule for 
pterosaurs.  I'm sure someone someday will chart the real rollercoaster graphs.
> 
>> Typos are to be avoided, but they don't ruin trees of sufficient
>> size
> 
> Warren 2007 fig. 10. Should I send you the pdf?


Please. 

And by "ruin" I mean the movement of major branches. And by sufficient size, I 
mean 50+ taxa and 150+ characters. 

David Peters