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

BCF, BADD, etc.



Another long-winded reply to JR Hutchinson, 95-12-04 11:39:14
EST:

<<A final, brief reply to G.O. (I feel the discussion is getting
a bit pedantic and longwinded for most of the list, so I'll agree
to hush now):>>

Likewise, a final response to the final reply. I hope.

<<>"The more characters used, the more accurate the cladogram
will be." How do you _know_ this?>>

<<Generally accepted by most cladists I know. Throwing out
characters a priori because you _think_ they might be homoplastic
or whatever is generally naughty.>>

This is actually an argument from authority: The existence of God
is "generally accepted" by as much as 90% of the American public,
but nevertheless others still consider it an open question. In
cladistics, the usual situation is that one worker's analysis is
overturned--not merely supplemented--by another worker's
subsequent analysis that uses more characters. And I need only
mention the struggles to be had with the geneticists and their
cladograms based on protein and DNA sequences, which are often
heavily at variance with strictly morphological analyses. If you can
somehow show that the addition of more characters to an analysis
_doesn't_ significantly change the cladogram's topology but merely
increases its resolution, then I will agree that your statement is
reasonable. (Some professional cladists try to do this, incidentally,
by presenting different cladograms of a particular group with various
characters removed to show how the resolution increases.)

<<What's good character choice? Read the dozens of publications
on it, e.g. Hennig, Carpenter and Currie's Dinosaur Systematics,
yada yada yada.>>

That was a rhetorical question. "Good character choice" is a
phrase that simply reeks of subjectivity.

<<>Ha! When is the last time you heard a cladist say, "Well, my
>cladogram says thus and so, but the stratigraphy says otherwise;
>so its back to the drawing board for me!">>

<<Just yesterday, when Michel Laurin (basal tetrapod cladistics
guy) admitted to me that his cladograms don't match the fossil
record perfectly, and hence are probably wrong in some places.
Good cladists (and there are plenty) don't make a cladogram and
then sit back and say, "OK, marsupials are done. On to
edentates." It's a process of refinement that really never
ends, but keeps getting closer to the truth.>>

Good man, Michel Laurin.

I wouldn't expect any cladogram to match the fossil record
_perfectly_. I will agree that the process never ends--but I
wonder whether it gets any closer to the truth(!). There are
billions of possible cladograms, but only one is the True
Cladogram. How can we possibly _know_ that we're converging on
the latter?

<<>I don't think there is any more validity to using parsimony to
sift through a bunch of cladograms _after_ they are generated
than there is in using parsimony to eliminate a bunch of
long-shot possibilities _before_ the cladograms are created.>>

<<You won't find many real (i.e. professional) cladists that
agree with you on that one.>>

Yeah. Probably not. But I'm interested in why they wouldn't.

<<>I _don't really care_ whether the archosaur phylogeny that my
methods and analysis have generated overturn BADD or not.>>

<<If you do care whether anyone believes you, then you'd better
try to convince them by their rules. Dinosaur phylogeny is
currently done by professionals via cladistics. If you're
inventing some new method to use in place of cladistics, you'd
better publish that first and subject it to peer review.>>

Hogwash. Cladists and professional paleontologists certainly
don't care whether I believe them. Why should I care whether they
believe me? In _Mesozoic Meanderings_ #2, I outline my method and
arrive at a phylogeny that is at variance with their current
phylogeny. I think cladistic methodology has some gaping holes in
it, and that my method, such as it is, yields more reasonable,
results, that is, a phylogeny that actually makes more sense from
a functional standpoint. Why should I use their methods and their
rules when I think they're flawed?

<<>No--if I play by the same rules, then I can expect to get much
>the same results, and we have made no progress.>>

<<I'd say you're wrong. If you replicate the same phylogeny, that
would support that phylogeny. If you use more characters and taxa
than previous studies, and don't omit included taxa, and come up
with a different phylogeny, _then_ you've got something, if your
methods are sound and the fossil record does not contradict it.>>

As I pointed out before, the entire fossil record of the most birdlike
theropods contradicts the BADD phylogeny, but this hasn't stopped
BADD paleontologists from going on their merry way.

<<>The objectivity of cladistics is compromised by the
unavoidable subjectivity of the data-gathering process>>

<<True in part, but name for me a 100% objective systematic
method. Right, there is none really. Name for me another method
that uses compiled data (ostensibly objectively measured) and
produces parsimonious phylogenies. Right, there is none really.
Good characters can be recognized by researchers trying to
replicate your study, objectivity is in the hands of the
researcher.>>

Cladists would certainly have you believe in the objectivity of
their method. I've read bunch of papers in _Cladistics_ and other
journals, and there is no question in their minds that they are
dealing with reality.

Look. If the cladistic methodology of accumulating everything
that could conceivably be considered a character into a character
matrix, running it through a brute-force tree-generating
algorithm, and presenting the result as a "phylogeny" worked,
then we would not see so much variation among the phylogenies.
Because one can pick subsets from a set of characters to support
practically any phylogeny (for example, by restricting my
characters to the fins and feet, I can create a phylogeny that
evolves fish from horses, via whales), there is never a guarantee
that the set of characters used in an analysis isn't merely a
subset of a still larger character set whose generated phylogeny
would be radically different. Indeed, since phylogenies of fossil
vertebrates are necessarily restricted to skeletal characters, we
may be missing the true picture of dinosaur phylogeny by a wide
mark--a picture that would be corrected if only we could include
the soft body parts and the genomes in the analysis.

<<>I have arrived at a phylogeny for archosaurs (excluding crocs
and Cenozoic birds, which I know very little about)>>

<<Ummm....that's not good. You have a paraphyletic phylogeny,
which is very naughty as you know. Might want to tidy that up a
bit; if you're going to do a revision of Archosauria, you'd
better include ALL of Archosauria. Yeah, it's a lot of work, but
if you're right, then I'll eat my words with a spork.>>

A "paraphyletic phylogeny"? What's that? Do you mean a
paraphyletic taxonomy? I simply accept current croc and bird
phylogenies and refer readers to the relevant works without
hassling with those groups. Did you think that my phylogeny
itself somehow excludes the higher crocs and the non-dinosaurian
birds? Not at all--they're there, just not in detail.

<<>I'm always interested in hearing from anyone and everyone who
disagrees with my phylogeny>>

<<Good, at least we're both being civilized here. I'd like to
add, G.O., that I enjoy your contributions to the list (as you
know, it wouldn't be the same without ya), and I'm not poo-pooing
BCF because I don't want to believe it, but because I see
fundamental flaws in it.>>

The flaws you may see are not fundamental to the _method_; they
may, however, be fundamental to _me_.