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

Re: SICB Report

DinoGeorge says:  ?The key problem with cladistic analysis is that it is
>inherently< unfalsifiable.?

That may be so for an individual cladogram, but the effectiveness of the
basic system may be tested by simulation.  Over Christmas, it occurred to me
that various imaginary scenarios of trees of virtual creatures could be run,
all busily evolving their characters at controllable rates, and the results
checked.  We could see under what frequencies of character reversals etc the
actual correct tree would be found, for any cladistic process.  With
increasing numbers of feasible simulations, the envelope of possibilities
will be increasingly adequately sampled ? this is called Monte Carlo
simulation, and in fact Peter Wagner (probably a friend of Chris Brochu?s)
advocated them in the ?Nature? discussion ?Is the fossil record adequate??
at the end of last year.

I imagine this was the closest ?Nature? felt able to get to the question
?Does cladistics work??.  I think the original website may be gone now ? it
was available by accessing ?Nature??s website ? but I?ve downloaded it.  It
makes very interesting reading.  (If anyone wants it electronically, I could
send it, or I might put it onto my website, if the demand is high. As
they've taken it down presumably to save space, I don't suppose they'd mind
someone else supplying it.)

On reading (some of) it, it is clear that such simulations are often done.
This should allow us to answer all our questions, though there will be a lot
of arguing, but it is possible to answer such questions as ?If these
characters are linked and often change together, will this be correctly
reflected in the cladogram??.

In fact, since most cladistic processes assume independent characters, they
will not be expected to handle dependent ones.  Of course, when it comes to
morphology, independence certainly cannot be relied upon.  Consider for
example the necessity for a bipedal creature to balance ? short-tailed forms
usually need ground contact further forward ? hence relatively immobile
femurs ? hence longer legs.  That?s one we?ve been reminded of recently.
And of course all the (other) flying/ex-flying features are not independent

But this is not a problem for those in the ?Is the fossil record adequate??
discussion.  Over and over again they carefully balance the various forms of
evidence: morphological-cladistic/molecular/stratigraphy, reflecting their
belief in each line of evidence having its own part to play, and carefully
gauging where each one starts to break down.

This is in contrast to the Padian & Chiappe Biol. Rev. 73 pp1-42 ?98
("expansion" of Sci. Am. paper last Feb) where despite admitting [p2] ?All
views must be regarded as subjective? preceded earlier by: ?Ours is not the
only possible viewpoint, nor the study one within integrative Biology [they
don?t mean theirs is not within I.B ! they mean ?not the only one
 within..?], and some workers have presented other conclusions or chosen
other approaches?? they continue ?[?] we base our approach on currently
accepted methods of comparative integrated biology, with a strong foundation
in [I would say an overwhelming bias towards?] phylogentic systematics?.
They end their conclusion emphasising: ??the need for further exploration,
new analyses and rigorous methodology??.

Rigorous cladistics is like rigorous 7-day weather forecasting.

The dino-bird cladists are not wrong in trying to use their version of
phylogentic systematics; they are wrong to place such a heavy reliance on
it, partly because they are using it outside its effective operating limits,
and partly because there are other methods not to be ignored.  Stratigraphy
is one, but there are many more.  Knowledge doesn?t come in 100% reliable
packets, and just as importantly knowledge is in too short supply for
unconsidered trifles to be discarded out of hand.

There are numerous ways to judge theories  (who would claim we have reached
the end in this field?).  When we compare palaeo scenarios against more
familiar ones, uncomfortable dissonances should not be ignored.  This is not
?intuitive hand-waving? as those whose intuition biases them unjustifiably
towards cladistics would say, it is instead a different kind of knowledge.
For example, a lineage of birds that gives rise to just one line of
flightless ground forms in 80mys clashes horribly with what we see today.
This is unbelievable, and surely statistically investigable.  The
Padian/Chiappe/Norell clade ignore this evidence.  Their supporters would
perhaps say only old fashioned people could think so unscientifically.

Likewise, failing to take account of differing explanatory abilities between
two theories simply throws away knowledge on how to think about theories.
Theories that are internally consistent and do not leave huge unxplained
holes usually turn out to be better.  A theory of solar eclipses that
explains them as a dragon temporarily swallowing the sun isn?t right just
because it provides an explanation ? but it?s still a lot better than a
theory of eclipses that doesn?t explain eclipses!  A theory that provides an
explanation of how creatures with characteristically large hands got them
isn?t true because of it, but it still scores points, on this round anyway,
over a theory of big-handed creatures that *doesn?t* explain how they got
big hands.  Likewise, it would be nice to have an explanation of why the
change to big-handedness waited 80mys before suddenly taking place.

Those who insist on relying on disprovability alone, yet still settle upon a
preferred theory before conclusive proof is in, are actually making use of
theory-choosing-methods they don?t admit to.  The only other method some
people often accept is parsimony, which leads them to place an undue
reliance on cladistics, though this does not honour parsimony as such, only
one (usually erroneous) instance of it, and usually to the exclusion of many
other instances of it.

The sixth contributor to ?Is the fossil record adequate? was Dr Peter Wagner
(Field Museum of Natural History, Chicago, USA) who wrote on:
?Why phylogenetic hypotheses need testing by stratigraphic data?

He warns about problems with cladistics: ?Unsurprisingly, simulations
confirm that parsimony analysis is less accurate as homoplasy increases and
global hierarchy decreases.?

Basically, if I interpret him right, he restates the often heard complaint
that the homologues cladisitcs is based on deteriorate in usefulness over
lengthening time spans.

I?m sorry, I can?t resist quoting this: ?Computer simulated evolution with
finite character states shows that parsimony often underestimates actual
evolution. This is true even when the shape of the tree produced by
parsimony analyses matches the known course of simulated evolution. Such
differences become more prominent when the simulations differ from the
initial assumptions used in the subsequent analyses. If the most likely
length of a tree is not the minimum length perhaps the abductive assumption
that "least is best" in phylogenetic analyses deserves re-evaluation.?

A final comment on the Padian & Chiappe Biol. Rev. paper : perhaps the
excuse of lack of space might have been used for not mentioning secondary
flightlessness in the Sci. Am. article, but in the fuller piece they find no
room for it in 42 pages!  Also, they managed to list  233 references they
considered more significant to the topic than ?Predatory Dinosaurs of the
World?.  And they didn't run out of space but left half a page blank at the
end of the article.  Presumably they will get an oportunity to meet Greg for
the first time at the "Ostrom" meet!



On ape and lizard climbing abilities...

"On a signal from Mohammed, one of the Pakistani workers caught hold of a
large Krow by its tail, body-slammed it against the high wall, and let go.
The Krow stuck to the wall.  The Pakistani climbed up the vertical Krow as
if it was a ladder.  I could see why the Krow was the burglar's best friend,
but still found it hard to imagine housebreaking with a giant lizard"

                                                                "Mr. Nice"
M. Marks

(No animals were harmed during the making of this posting.)