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

Re: The actual running Archie paper...(this is now even longer I'm afraid!)

Thanks for your careful and considered reply, Matthew:

--Original Message-- From: Matthew Bonnan <Z966341@wpo.cso.niu.edu>:
Saturday, May 08, 1999 05:34 AM


>Okay.  I am looking at pg. 217 in Greg Paul's Predatory Dinos book.
>First off, do we know for sure that the supracoracoideus muscle in
>Archaeopteryx actually formed a tendon, similar to that of a living
>bird, that passed through a groove on the scapulacoracoid before
>insertion onto the humerus,

There is a prong on the scapulocoracoid of _Ax_, in the equivalent place to
that in moderns.  It is reasonable to guess that this does the same job, and
surely also reasonable to guess that where the s.c. makes that massive
change in direction it takes the form of a tendon.  However, it wouldn't
really matter if it was muscle at that point.

>   ...or would have followed a pattern similar
>to what was likey the predatory dinosaur condition, as shown by Paul
>in the Coelophysis drawing?

In the Coelophysis pattern, the muscle still pulls forward and up.  It is,
to say the least, unparsimonious to consider the _Ax_ prong not to function
as it does in moderns.

>Second, I'm not sure I understand your pulley argument.  Are you
>trying to argue that Archaeopteryx could fly because it had a powerful
>supracoracoideus even if it did not pass through a triosseal
>canal-like structure?

It had a fairly powerful s.c.  It passed around a pulley, though not through
a tunnel, but this latter detail is not crucial for our purposes at the

>... Or that the pulley mechanism is a "red-herring"
>because even if Archaeopteryx possessed one (if it did) its
>supracoracoideus would be too small to have made it useful as a flight

Its s.c. was big enough to be useful; the red herring is that not having a
tunnel is irrelevant to whether it had a pulley or a useful s.c.

>Also, to my knowledge, the triceps tendon (is that what you mean by
>"outside elbow tendon") is NOT a pulley.

Ah - right!  Thanks for pointing that out!  Now I come to think of it, I
think I remember that myself.  I have to say here "Perhaps that was a bad
example" - and I know people will take that as a collapse in my entire
argument, but just substitute any true pulley system of your choice for a
proper example.

>I bring this up because you are correct, a tendon need not pass
>through a tunnel to be a pulley, but pulleys have special conditions,
>one being that they change the angle at which a muscle inserts to gain
>some sort of mechanical advantage, so you see that tricpes tendon
>cannot be a pulley system.  This is why (as I assume you know) the
>supracoracoideus passes through the triosseal canal to insert on the
>deltopectoral crest of the humerus -- it provides a better angle of
>insertion and mechanical advantage: it's a pulley.

As I recall, "mechanical advantage" in the strict sense applies to some sort
of gearing - as in the Western Differential system - in other words you pull
the rope going into the system eg twice as far as the rope coming out of the
system moves, but the rope coming out of the system pulls twice as hard.

However, if you only want the rope to go round a corner, you might loosely
term that as some kind of mechanical advantage.  Anyway, it's still a
pulley.  Most muscles, via a pulley or not, work the other way - the final
thing they affect moves faster than the muscle itself but with less force.
That certainly applies to the triosseal canal in moderns, since the wing
moves much faster than the s.c. muscle.  That system does not show
mechanical advantage in the strict sense.

>So you know, I took the "dare" and went to your web site.

Hey hey!  Good on yer!

>... I do not
>wish to debate with you point by point your argument against the
>ground-up hypothesis on bird origins,

That's one reason I set up the site, so as not to have to reiterate
everything over and over again.

> ...but I am curious as to why you
>suggest we "disbelieve" cladistics.  While some cladistical studies no
>doubt abuse the power of the hypothesis generator, surely you don't
>suggest we toss out cladistics entirely?

If they abuse the power of the hypothesis generator, we should certainly
toss out the unjustified importance given to it.

>Many folks I know spend so
>much time and energy to make sure their characters are meaningful,
>non-repetitive, and testable that it is hardly the
>dump-everything-in-and-see-what-we-get approach you seem to think it

Have we yet reached the point when as much care and diligence has been
expended on cladistics as was expended on enhancements to the Aristotlean
model of the universe? ;-)

>But, let's say you are right -- cladistics is ultimately misleading
>and poorly designed for fossil vertebrates.  Fine.

Well, its quite a good idea in densely evidenced domains.  In those cases it
doesn't really do much more than save manual labour though.

> What are you going to replace that methodology with?

What am I going to replace poking myself in the eye with a sharp stick with,
now that I've decided it's not a good idea? :-)

>Which former method of phylogenetic
>hypothesis can you suggest that has the same testablility,
>falsification, and predictive power of current cladistical

I'm a creator, not a follower.  (Though I certainly wish I wasn't!)

>Or, can you suggest to us a method that encompasses the
>strong points of cladistics but blows away the parts cladistics cannot

Using common sense for one thing.  Cladistics is exactly the kind of thing I
might have devised myself, but had I done so I would have described it as  -
well, a hypothesis generator, or a labour saving device, or perhaps just
some kind of toy.  I would have expected it to produce certain dodgy
outputs, and would have used other evidence to help me mark out which of
them to suspect.

One thing I have suggested is not throwing away everything but cladistic

Another thing I have suggested is to use homomorphisms one has great
confidence in - for example uncinate processes on the ribs which have
proven, since the Cretaceous, to be incredibly stable in dinobirds, and
discard dodgy ones, which show evidence for over-variability.  Then
progressively break down the specimens into smaller and smaller groups,
using only features we have reason to trust.

I would also advocate a more judicious approach to general principles, and
less awe in the face of printed calculations.  There are very few things we
can can rely on absolutely; one of them however is that over-reliance on
simple principles is a mistake.

>Until you or someone else can do that, cladistics is the best
>thing we have.

I'm as likely to do it as anyone.  Not that anyone will listen to me though!

>And I by no means suggest that it will stay forever or
>never be replaced by better methodology -- all I suggest is that you
>cannot dismantle cladistics without a replacement.  If you don't like
>cladistics, suggest something else.

It is not a general principle that everything that is bad needs replacing.
(See the "sharp stick" argument above.)

I do however have an alternative.

All indirect knowledge involves generalisation; the problem is of course,
which of the umpteen possibilities is anything a generalisation of?

For example, when guessing pack hunting in Deinonychus, shall we say it is
an example of a reptile?...in which case we might generalise from extant
examples.  Or is it an example of a plains carnivore predating much larger
prey? ...in which case we might invoke African hunting dogs as the example.
Or is it an example of a bird?  Of a flightless bird?  Of a dinosaurian
predator?  Of an archosaurian predator?

The same plurality of possibilities applies to para-theories:  should we
invoke parsimony?  Should we invoke the rule that parsimony shouldn't be
relied upon in an area in which our knowledge is sparse?  Should we even
rely more on para-theories, or on more specific rules such as those in the
previous paragraph?

If we go for parsimony, should we minimise the number of individual changes?
Or maybe the number of niches or roles the creature was adapting to? (eg
flight, or the elusive "activity X" that give rise to feathers, wings etc
according to the popular view).  The latter might sometimes be more
reliable, but if it were, it would be harder to justify any particular
quantitative treatment for it.  Another possible application of parsimony is
of course in the stratigraphic sense - assume age indicates position in the
tree unless very good evidence contraindicates this.

Another guideline might be: If some adaptation has not been made by a
lineage for 100 mys, and then suddenly is made, be wary of any theories that
say "it just happened"; prefer those that at least offer an explanation.

Another might be: If some vast suite of adaptations suddenly appears
simultaneously, all of which are astonishingly perfectly suited in great
detail to some particular role, again be very wary indeed of theories that
say "coincidence: they just happened for reasons unknown".  Statistical
justifications for this guiding rule might conceivably drawn up - inadequate
to satisfy a disputant no doubt, but perhaps useful for believers.

I'll stop giving possible rules now; obviously they have been drawn from
familiar topics, but they are intended to illustrate a point:

We do not live in a neat mathematical simulation inside a computer, where
everything is certain.  We live in the real world where there are in
general, no rules except those we make up, but even those are liable to
correction at any time.  Our best guesses are best inspired by a vast expert
system than a fixed list of "laws".  The search for better rules for
generalisation is never ending.

>You mention Occam's razor and its supposed misuse in paleontology
>because of the possibility of character state reversals.  Yes, you are
>correct, we should be careful and understand the fossils we are
>dealing with well.  However, you also explain that the first "flaw" of
>cladistics theory is "How do you decide what feature(s) to choose when
>defining you clade?' and then suggest this is all arbitrary and
>ultimately very theoretical with little basis in reality.
>Okay.  So let's examine the dinosaur-bird origin.  Well, what
>characters should we pick?  We could pick any!  But, of course, some
>characters are more valuable than others in providing information.  We
>could count every bone molecule and make that a character, but that
>would probably not be too informative, nor would measuring the length
>of each animal and plotting them out -- characters of size.
>I don't think you'd disagree that birds are archosaurs?

I realise I haven't made my theory clear enough, and you think I am a
Feduccian.  Though I sympathise a lot with him, I don't when it comes to
details of his bird evolution beliefs.  I'll definitely go back to the
website and make a few changes.  I think that birds come from good theropod
stock, it's just that the Cretaceous manis (and a few others) came from
Archaeopteryx and not one of the reverse possibilities.

Thanks again for taking the trouble,