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

Re: hypotheses, science 'n' stuff



Some more comments now that I jettisoned my relatives visiting for Memorial
Day...

Ralph E. Chapman
Applied Morphometrics Laboratory
National Museum of Natural history
ADP, EG-15  NHB, 10th & Constitution, NW
Smithsonian Institution
Washington, DC 20560-0136
(202) 786-2293, Fax: (202) 357-4122
Chapman.Ralph@nmnh.si.edu

>>> <Dinogeorge@aol.com> 05/30/00 11:10AM >>>
In a message dated 5/30/00 9:49:27 AM EST, chris.lavers@nottingham.ac.uk 
writes:

<< We always have to justify our
 belief in any particular epistemological handhold - even logic - with
 reference to something else (I believe in x because...). This means that
in
 justifying our belief in any methodology, we must necessarily enter a
state
 of infinite regress, constantly justifying our belief in one thing with
 reference to a more fundamental thing, or make a leap of faith (I believe
 in logic, so there!). A leap of faith is a leap of faith whether the
 landing area is logic, maths, or god, so all approaches to knowledge must
 be indefensible. WWBIII's reasoning also has the charming feature of
 refuting itself,  which just makes it all the more attractive, IMHO. >>

This is a wonderful summary of the very point I was trying to make in a few

of my recent emails but obviously failed to do satisfactorily.


>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

We can argue about the basis of science according to philosophers, who I
suspect rely wholly on the results of science whenever they try to do
anything in their lives like travel or get medical care, etc.  The ideas
here basically go in and out of favor based on the philosophers who are in
important academic positions, and in cycles. The basal scientific approach
of hypothesis proposal and testing pretty much is unaffected by this and
continues on and, happily, has just been getting more common. However,
regardless of the basal structure or belief structure, the methods followed
by scientists still are totally different than those of religions in that
the latter rely on belief in the word as handed down, and the former relies
on an evolving set of data that can be and are tested following
well-established procedures. You can certainly argue the merits of both, or
even find them not contradictory in the gestalt of life, which I sorta do,
but they are different. One acts as hostile or at least very resistant to
dissention, the other rewards it, as long as significant data accompany the
dissent. They are also irrelevant to the main discussion we were having when
George brought up this little side-track.

What started this whole process was the bold assertion by George that
paleontologists cannot  do science as scientists in other areas can who can
see their organisms live. As such, he asserted, we were relegated as only
doing just so stories. This is a totaly different question. First a word
from our sponsor...

Just so stories are fun, they are the start of the process of doing
science. If you go to the amusement park and get in line for a ride, the
ride is the show and the line usually has various things to keep you
occupied - the preshow. Actually, real artists like Disney or Universal
usually have the pre pre pre show and the pre pre show as well, along with
some post show and the shop for the show. Well, just-so-stories are the pre
show, where you sit back consider the data you have and make up a story
about how something happened. When you get to the, makes good sense to me
stage, you have your just so story. In the old days, lots of these were
basically published in that form and they traveled with strength based on
the authority figure who proposed them. This has an interesting irony to the
Jackson email about philosophers which indicated an unnamed philosopher of
top authority recommended by the best philosopher (authority #2) from the
best philosophical institution (authority #3) said cladistics is crap. Even
authorities are full of crap sometimes, even this one (easy home run using
this phrase against me, George), and we all need be wary of the tyranny of
authorities when we do science, even as we respect the knowledge that
authorities have. We just can't allow authorities to get lazy and not do the
work.

However, just so stories only perk the interest. To actually move to the
science, you have to then develop the hypotheses that are relevant and try
and test them to see if they can be destroyed. The latter part is the stuff
of science and, despite the protests of George, paleontologists do this as
offten or more than those in the neontological sciences. When someone says
that paleontologists do just so stories only, it is an insult and is
inaccurate. Suggesting George just find a new name for them is silly.
Simply, George seems just to have no feel for proposing and testing
hypotheses.

This also came through with his first statement about protecting his
hypotheses from assault - making sure they win out eventually. Totally wrong
scientifically, as was strongly pointed out, and all the crawfishing in the
world can't stop this first statement from being another strong indicator of
a problem here. Make others who argue against a proposed hypthesis do it
well, sure, but you have to be as tough on your own proposal as new evidence
comes to life, and it always does. I suspect George does this even more than
he thinks he does and I'll let him off the hook and say we all have problems
at times letting go of our babies. With hypotheses as babies is better to be
sea turtles than raptors (the birds, that is).

I'm trying to do the right job with pachy domes which gave me problems with
a newspaperman one time who was interviewing me. Horner (with Mark Goodwin)
gave a talk suggesting that pachy dome microstructure precluded head
banging. I found the evidence here far from compelling and, in fact, the
domes just looked basically pathological to me, which you would expect from
head-banging. They were doing their job trying to refute Galton's hypothesis
and subsequently have lots more data to present that I'm looking forward to
seeing. I was doing my job in not being convinced of their refutation
because, frankly, it was not there. The news guy asked me about Jack's talk
and I said I what I thought. He then said "So you think they banged heads
then." Not a question but a statement. I said to him, "No, I saw no reason
from the talk to refute that idea but that I had lots of work of my own to
do before I was totally convinced one way or the other. Basically, I said
Galton's hypothesis was not disproven by the talk, in my opinion. The
journalist got really huffy that, after 5 minutes of additional argument, I
would not just say that pachys banged their heads. He lost the distinction -
or never had them in his mind in the first place - about how science is
done, how hypotheses are proposed and sometimes rejected.

The arguments that various individuals have made show that we indeed can
frame arguments in paleontology scientifically in the same way as any other
science can. Hypotheses are testable and potentially disproved. They are not
provable. They can only build up such a large and complex supporting body of
data that they will tend to be very robust to smaller perturbations in the
data itself - as with ornithischian herbivory. It's nailed because it would
take such an exceptional body of data to refute it right now that
evolutionary biologists can accept it as being as probable as we can get,
despite being a type 3 inference. Evolution is similar but in spades - Leigh
Van Valen one time suggested that it was as solid as the theory that there
are molecules, and he was right on. Creationists have never noticed that one
badly described and misrepresented statement that they see as countering it
will never work. May work for their religion, but that is a different ball
game, except possibly for George.

Despite suggestions to the contrary, we cannot "prove" that diseases are
caused by the things we think cause them. We can set up tests that are not
disproven. We can build up such a large, robust and complex body of data
that the hypothesis is compelling and treat the disease following this. The
whole HIV does not cause Aids lousy argument that keeps living on  (also
mentioned by Mickey) is a strong showing of this, even though the data are
overwhelmingly against it, there are always some that will say the
connection is not proven. So the original argument by George, frankly, is
totally incorrect, even if only within what some consider the flawed context
of science in general. Paleo is no more baloney than physics. Paleo is
Filet.

I remember the posters on the wall of opening day at the movie Close
Encounters that said "There is no scientific evidence that UFO's don't
exist". Guess we'll have to keep looking...

Jackson urged George to stay the course against my cladistic ravings. The
problem is that I did not actually claim cladistics was the be all, did not
even mention cladistics. I just implied there were procedures - here
scientific procedures - for evaluating data. The main thing is that whatever
you do, whatever algorthm you employ, the procedure must have explicit
methodology and be repeatable. You can use whatever algorithms you want but
only a fool would argue against using procedures that are repeatable and
explicit in what you do. So don't put algorithms in my mouth. However, I
can't guarantee that, even if you are explicit, that other scientists will
be compelled to find you compelling. The vast majority have accepted certain
methodologies and it's up to others to come up with new ones that have the
strength of being explicit, quantitative and repeatable, but have better
behavior than standard cladistic ones. Would love to see it because, as I've
said, most still see problems with all the algorithms for this.

Anyway, we can talk about most phylogenetic analyses as being an
optimization procedure that can then be folded into a more structured
scientific approach, just as statistics are used as part of all such testing
and hypothesis building, but I think we've pretty much run our course -
paleontology is a science that relies on scientific procedures, only some of
its proponents still must rely on just so stories.


Ralph Chapman