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

Re: (Paleognath monophyly)

John Pourtless wrote-

But it doesn't seem to be a reversal, parsimoniously speaking. Of course, you argue against the strict application of parsimony, but without it, you could call pretty much any character that appears outside your ingroup a reversal. "Toothlessness is not a good synapomorphy of Neornithes because it could be a neotenic reversal to the primitive state exhibited in Confuciusornis and Gobipteryx. Sure, given parsimony, these are more likely to be convergences (as all other enant-grade birds, Jibeinia, yanornithids, hesperornithines and Ichthyornis are toothed), but we can't trust parsimony."

I did not say that parsimony could not be trusted, I said that blind adherence to it, especially if there are alternative phylogenies which are just as possible, is not to be trusted. Character weighing is inevitable in phylogenetic reconstruction, and it will never be turned into a completely objective discipline. I refer you to Ghiselin's 1984 critique, since he said it far better than I have.

Alternative phylogenies that are just as possible according to what criterion? There are almost always other topologies that seem plausible from a subjective standpoint. But we must have an objective method for determining which hypothesis is better supported, or else phylogeny ceases to be science and starts being a competition of personal preferences. Character weighting in morphological analyses may be inevitable (once we understand the genetic mechanisms behind various types of transformations), but we're nowhere near that point yet, if it ever occurs. Until that time, such weighting is subjective, and thus not science.
Could you provide the citation for Ghiselin 1984?

Cranial pneumaticity caudal to the quadrate and the inflated alaparasphenoid are the strongest of the characters listed, and the only two which Cracraft & Clarke produced that I would call "unambiguous," however, two such characters do not amount to an overwhelming advantage to the paleognaths-are-holophyletic concept, and at the very least, it that seems that multiple possible phylogenies of Paleognathae are equally viable.

Well, we'd need some method of weighing alternative hypotheses. If you don't like parsimony, what objective method do you suggest?

I feel that the formulation of auxiliary hypotheses, (again see Ghiselin 1984), as Popper defined them. To quote Ghiselin, "There need be nothing ad hoc in phylogenetics about invoking stratigrapjy, biogeography, genetics, embryology, or ecology."

I agree regarding the last three, though the first two seem inappropriate to use for vertebrates at least, given the scarcity of their remains. I suppose I'll need to read Ghiselin to see how auxiliary hypotheses come into play. If you mean we can weigh alternative hypotheses by seeing how they compare to hypotheses formed from other data types (morphology vs. genetics, etc.), there still needs to be a criterion for weighing these characters against each other. And why not just combine all types of data into one analysis, as is commonly done today? But I may be interpreting what you said incorrectly.

> Paleognathae is entirely polyphyletic, and secondarily derived from within
> Neognathae,

>From _where_ within Neognathae?

As mentioned tinamous seem to be more closely related to the gallinaceous birds than they are to other ratites. 

That covers tinamous.  Do you propose this topology then?
|--Ratites
`--+--Neoaves
   `--+--Anseriformes
       `--+--Galliformes
           `--Tinamiformes


I would propose the following topology

Neognathae
    Galliformes
    Tinamiformes

This places the two in an unresolved dichotomy, which I think is the most conservative placement. If tinamous are related to Galliformes, I would argue that tinamous are derived from within Galliformes, and not a sister-group thereof. Again though I am opting for a conservative approach, until more data becomes available.

Well, that's sort of a topology. :-) I don't think "unresolved dichotomies" exist, by definition. But okay, you think tinamous are more likely to be galliformes. Maybe tinamous can be added to Dyke et al.'s (2002) morphological analysis of galliformes, if they aren't already an outgroup (if they are, we can make them an ingroup). I'll have to check my copy of the paper when I get back on campus Sunday.
BTW, one major advantage cladistics has is that it forces people to try to find explicit relationships. No longer do we have well known taxa placed somewhere vague "until more data becomes available". Instead, they get included in analyses and will definitely go _somewhere_, based on characters that can then be checked and compared to competing phylogenies. Without it, you end up with most taxa coming from vague unknown ancestors of uncertain relationship to each other. We shouldn't have to wait until more data becomes available to place tinamous in a cladogram- we've been studying complete live specimens for over a century. Sure, currently unknown data may change what we think now, but we don't need to have the right topology, just the best given the data we have.

Whatever you think of Feduccia's work on the origin of birds, he has done important and accurate research in his career on various matters pertaining to the phylogeny of birds. It is one thing to disagree with a man's work in one area, but disagreeing with him entirely in everything he does (which some, though not necessarily you, seem to do) because of his work in another field, is unwarranted. For instance, I think Gauthier's restriction of the term Aves to a crown-clade is rather useless, but that does not lead me to dismiss all of Gauthier's work.

Admittedly, I haven't read any of Feduccia's scientific work on neornithines. I read his popular 1996 book several years back, and basically everything controversial about neornithines I can recall from it is no longer believed today (Presbyornis being a duck-flamingo link; shorebirds being the basal neornithine grade; parrots being derived from cuckoos). He might have gotten something right, I'd need to read more of his work to find out. But given his methodologies regarding the dino-bird debate (think of evolutionary scenario first, try to work taxa into said scenario, use only a few characters to decide the phylogeny, write off anything else as convergences), I don't see why his neornithine methodologies would be any different, and I don't think they are at all good for finding phylogenies. I'm not saying everything he does is potentially flawed, just the phylogenetic aspects. He may do great morphological description or paleoecological interpretation for all I know.

And what pedal characters are there to support Oviraptorosauria? None I know of, and they exhibit quite a range of morphologies. Arctometatarsalian caenagnathids with fused metatarsi, the elongate subartometatarsalian pes of Caudipteryx, the robust non-arctometatarsalian pes of oviraptorids, with a lateral digit reduced in Conchoraptor, the extremely gracile fused arctometatarsalian pes of Avimimus with metatarsal V fused to it and no digit I. We should presume the pedal morphology of oviraptorosaurs is a funtion of common ancestry because of the non-pedal characters that unite them. Similarly, we should presume the pelvic morphology of ratites is a function of common ancestry because of the non-pelvic characters that unite them. Disparity has no place in phylogenetics unless you can show it's also synapomorphy (so if some ratites shared pelvic characters with some neognaths, you'd have an argument). Then we'd have to weigh these hypotheses using whichever method you prefer.

That argument relies on the validity of the alleged cranial synapomorphies, which is still contested. 

Regardless, the disparity in pelvic and palatal morphology is irrelevent.

Ostriches are paleognaths and are perhaps derived from within the lithornithithid grade. The closest link between these two taxa is _Palaeotis weigelti_, but the first _Struthio_ fossils do not appear until the Miocene.

Here, it's very important what type of grade Houde thought lithornithids were. Paraphyletic to Neornithes, or paraphyletic to Ratites? I suppose I'll find out come Monday.

Houde argued, IIRC, that they were paraphyletic to ratites.

So, assuming you're supporting Houde's interpretation, you're arguing for a diphyletic Paleognathae, with tinamous probably being galliformes, but ratites being monophyletic (assuming fossil taxa like lithornithids can be considered ratites) and basal to other living birds.

As for my continued participation, I see no reason why I should not do so, though I am deeply distressed by the strong assertion (or at least implication) made in this list that cladistics is essentially the only scientific way of reconstructing phylogeny and everything else is pseudoscientic nonsense. Cladistic zealots are as frightening as any other zealots.

What other scientific ways are there other than joining taxa based on synapomorphies? Cladistics doesn't equal "unweighted analyses that use parsimony as their only criterion to choose topologies". But even accepting the latter definition, how can weighting morphological characters be objective? Where's the literature saying "character type A is prone to homoplasy, so should be weighted only .7", or "character change Y requires X amount of mutations compared to change Z, so should be weighted 1.4 times Z"? If you don't have this kind of data, you'll only end up with conflicts that can't be objectively resolved. "The number of secondaries is more important! No, the carotid structure is! Why? Because we each say so." And that's not science.

Mickey Mortimer

_________________________________________________________________
Let the new MSN Premium Internet Software make the most of your high-speed experience. http://join.msn.com/?pgmarket=en-us&page=byoa/prem&ST=1