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

Re: Just So stories and science

Matt, I'm a bit reluctant to respond to your post because I know that you
have work to be doing.  However, this is a discussion improving by
accumulation, without disagreement, so I'm going to assume that you can
enjoy it as a diversion when possible.

There's an apparent contradiction between two of your comments I'd like to
examine.  First you said:
<But as all good scientists realize, the easiest person to fool is yourself.
That's where peer review comes in.  You present your data and observations,
explain what you think is going on, propose a testable hypothesis, and throw
it to the wolves.  If your data checks and your observations are repeatable,
and others come to similar conclusions, perhaps your hypothesis has some
If not, chinks in the "armor" show up quickly.>
Then later:
<>If I share a bias toward a certain type explanation with everyone else I
have to persuade, then I'm going to have an easier time selling my
Unfortunately, yes.  However, it is the evidence that wins converts to a
particular hypothesis or theory about dinosaurs.>
The peer review process would break down if the peers share the same bias
toward a particular type of explanation, say adaptation.  The reviewers
might not perform the 'test' of looking for an alternative explanation which
also includes all the available evidence.  The need to consider this
potential problem arises, I'd argue, because the less data available, the
more potentially correct hypotheses could be devised.

You did discuss a potential solution to this problem when you said:
<I suppose one could defer to parsimony, for instance.  If there are two
competing hypotheses about, say, theropod evolution, one might choose the
hypothesis that requires the least amount of "steps."  From a functional
standpoint, although two competing functional scenarios might appear to have
equal evidence, the more simple of the two hypotheses here would probably be
preferred pending further evidence.>
And Gould and Lewontin use the term 'Just-So Story' to describe adaptation
hypotheses so elaborate that significant tests are just about impossible.
So, Occam's Razor, the simpler explanation is better, is a good way to
choose between hypotheses with great difference in complexity, and parsimony
is a good way to choose between two comparatively simple explanations.
That said, I don't think you would argue that parsimony is correct at all
times and in all situations.  In fact, as I understand it, other criteria
besides parsimony are also being used to decide between/among hypotheses.

The assertion we're considering can be broadly stated as 'Paleontological
hypotheses usually remain more tentative than hypotheses available to
non-historical sciences.'
At the extreme, it could be argued that most paleontological hypotheses are
in the situation you described for warm/cold blooded dinos:
<In this case, perhaps the best thing to do is keep throwing everything we
have at it but to try and remain as neutral to the results as possible.  In
other words, sometimes in science we have to wait and be patient, and many
times this is very difficult! =) >
This extreme position would substitute 'usually in paleontology' for
in science' in your discussion.  Somebody could urge 'always in
but I won't!

One useful way to escape this extreme version is to identify examples of
paleontological hypotheses which can be considered solid, an assumption
for future hypotheses.  You did mention the example of evolution, but that
theory certainly does not rest on paleontological evidence alone.
I'm not sure that cladistics should be included.  As it is used, doesn't it
on assumptions about what is usually true about relationships among animals
logical rules such as parsimony which are usually correct?  Here's where the
discussion about alternative explanations and logical necessity comes in.
If I
know that logical alternatives to my hypothesis are possible, isn't my
a rickety structure on which to build new hypotheses?

One promising example is the eye adaptation hypothesis.  If Pinker and Bloom
correct that such elaborate function must have happened as a result of
(though you're right, they should have noted the eye probably began as an
their definition needs clarification), then we have a solution which does
not allow
for alternative explanations.  Having urged that wings/feathers/flight are a
situation in the discussion about progress, I'm really rooting for this one.
Glad you're reviewing it!
If this works, then we can say that eyes can produce less tentative
hypotheses.  Not
by themselves of course; different populations of the same or another
species can
develop elaborate eyes.  But given enough samples with data about time and
geography, couldn't tying relationship to elaborate and obvious function
provide a less
tentative hypothesis?

You're very right when you said:
 <As human beings we want to know why and  want answers, and that's what
parts of paleontology both exciting and frustrating.>

Appreciative regards,