Re: Kerberosaurus manakini

[David Marjanovic <david.marjanovic@gmx.at>]

----- Original Message -----
From: "Mike Milbocker" <mmilbocker@psdllc.com>
Sent: Tuesday, June 22, 2004 2:15 PM

> > Phylogenetics is a useful tool, but is not the sine quod non of
systematic
> > paleontology. Unless a new specimen, or a redescribed specimen, suggests
a
> > novel phylogenetic hypothesis that figures prominently in the paper,
authors
> > should not be compelled produce a tree.

They _do_ propose a totally novel phylogenetic hypothesis -- the first
_ever_ that includes the new specimen. :-) I'm interested in the
phylogenetic place of a newly described species, so I definitely want its
description to contain a phylogenetic tree, means, a cladogram -- if, of
course, the specimen isn't _too_ incomplete.

> > There should be a place in this
> > science for a good old-fashioned comparison and discussion.

Comparison is part of coding character states for a cladogram. A cladogram
without discussion ("why is the tree so weird") isn't worth much either.

> > Very often, an informed, intelligent description and comparison says
> > as much as, often more than, a phylogenetic analysis.

It's still good to have it illustrated. And if one already has done the
description and comparison, the coding becomes a matter of half an hour
(assuming a small matrix).

> The application of systematics goes astray, in my
> opinion, when people use it to "prove" other competing theories,
hypotheses
> false. Bottom line, systematics is merely a data ordering tool made
> quasi-scientific by the often misunderstood principle of parsimony.

"Data ordering tool"? The goal is not at all to order data. The goal is to
produce a phylogenetic _hypothesis_. Hypothesis emphasized because a
traditional genealogical tree is a speculation that is often not
falsifiable -- cladograms are always falsifiable.

> I see no reason why a specialist should be forced to
> justify his conclusions in light of a cladistic result.

When those conclusions are about phylogeny, then what better method is
there?

> Further, no one way of ordering data can be
> proven superior to another without empiricle evidence.

Isn't a data matrix empirical evidence? :-)

> Researchers who drop characters (which they should note) to illustrate
> a way of viewing the data consistent with their "good old-fashioned
> comparison" work may ultimately be vindicated by the fossil record.

Yes, but they'd be right by coincidence, for the wrong reasons.

> Ultimately, the value of what's written in a scientific paper
> depends more on the intellectual honesty of the authors
> than any supposed objectivity imposed by cladistics.

Sure you can screw it up. But cladistics forces one to greater degrees of
intellectual honesty than other... than the _lack_ of a method. :-) It
ensures that the authors make many of their assumptions explicit, and that
the outcome is the simplest way to explain the data. When the data were
manipulated, this is another problem.

> The relations among taxa described in cladistics is not
> a scientific fact, only the states of the characters are.

Erm... bingo. Nobody claimed they were facts. The very point is that they
are hypotheses -- and not speculations anymore.

> When I was a youth studying astrophysics at MIT we used to joke that the
> value of the Hubble Constant could be plotted as a function of time,
that's
> how I feel about the validity of any one cladistic result.

I get the joke, but note that this function wasn't linear. It started to
oscillate and then to converge. :-)