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

Re: Feathers are not magical things...

I think I understand:  it is possible for a 'scientific' idea to be proven
factually wrong and for an idea which cannot be challenged directly to be
'unscientific'.  I think Jura was arguing that the term 'scientific' should
be reserved for statements proven by direct observation or so many indirect
observations as to become a 'theory' like evolution or relativity.

Returning to your usage, here's an example of being scientific, but wrong
(in the sense of factually
<Was the original study "flawed"?  Perhaps in a very strict sense of the
term.  However, we as mere mortals cannot operate in a sphere of absolute
knowledge: all we can do is sample from the whole universe of data.  Our
models (regression lines, cladograms, etc.) are only as good as the data
that is put into them.
[By the way, there also can be problems with which off-the-shelf model, if
any, applies in a given situation.]
As new data is added, the conclusions might change.  They might change just
a tiny little bit, and so the essential original
conclusions remain.  They may change a LOT, and so refute (for the moment)
the original conclusions.>

And after a number of examples of ideas identified as not demonstrably
wrong, but
<The short answer: no, I would not be "outright wrong"; however, I would be
unscientific.  (These are different types of things).  It is, ultimately, a
burden of proof issue.>

Okay, I think I see it.  A scientific idea is ' a hypothesis backed by a
self-consistent logical argument employing certain principles'.  You're
that no alternative argument is possible given these principles, and
therefore a
proposed alternative must be so poorly supported as not to meet the
definition of a scientific hypothesis.  Also, that denying such a well
idea must be subjective rather than analytic.

You referred to one of the principles when you said:
<In the sciences we choose the simplest explanation possible, since that
criterion requires the fewest additional assumptions.>
However, you are not arguing for just characters and simplicity.  (I didn't
realize this in a past discussion.)  You do include:
<*Models of adaptation and functional morphology
*Patterns of biogeography
*Evolutionary tempos and modes (e.g., Punctuated Equilibrium vs. Phyletic
Gradualism; Red Queen matters, etc.)
*And just about anything else that gets published in journals like
Paleobiology and Journal of Experimental Zoology and the like.>
So, anyone with a character list and PAUP would not necessarily be able to
produce a 'scientific' cladogram.  There are certain conclusions which can
be dealt with by simplicity, such as the examples that you gave, but
producing a hypothesis about a whole set of relationships involves
application of a range of other knowledge.
Your 'burden of proof issue' involves a whole complex of ideas.

So, a scientific argument about relationships includes:
- a number of shared characters judged sufficient to indicate that certain
animals are
related and few enough shared characters in animals outside the group that
animals can be excluded,
- expectation that unknown characters will be consistent with others in the
group, as in feathered velociraptors,
- expectation that evolutionary change in the group will be gradual and
(after all, if a large number of brand new characters suddenly appeared in
the lineage,
 the character similarity screen might fail to identify the animals as part
of the group),
- survival of the logically determined hypothesis after application of other
constructs, including models of adaptation and functional morphology,
patterns of
biogeography and the other examples given above.

As you've probably already seen, the issue that bothers me is that the
pre-screening of hypotheses may not be consistent with the later applied
constructs.  Suppose you could establish that PAUP gave a lower weight to
punctuated equilibrium scenarios than to more gradualist models.  Would that
invalidate the algorithm?
I realize that this must have occurred to other people; I'm asking because I
solve it myself.

Because of something that happened recently, let me pause long enough to say
I'm glad the application of principles including but not limited to
simplicity isn't simple.
If someone wants to criticize conclusions about pathological gambling
prevalence, they should know about SOGS, DSM IV, survey methodology
including weighting by demographics and the standard error of
subpopulations, false positives/negatives from lifestyle assumptions built
into the instrument, Winter in Minnesota (pun), twins studies, comorbidity,
incremental costs...
On the other hand, looking within arguments for ambiguities, unnoted
assumptions, and arbitrariness is welcome.
Okay, end of soliloquy.

Finally, you observe:
<And this is where the difference between Platonic and Darwinian models come
into play.  Platonic models of life says "the type is real; variations from
the type represent oddities".  The Darwinian model says "the type is a human
construct; variations (both within species and down along lineages, bringing
with them their historical baggage) are the reality".>
But you also said:
<If you were drawing a "typical" dinosaur (whatever that means), go with
In that sense, aren't feathers an 'oddity' among dinosaurs?
And if there weren't strong similarities within lineages, wouldn't
cladistics, etc. be
impossible?  In fact, that consistency is the basis for your point about
Not disagreeing with the basic point about not being blinded by archetypes,
isn't it a bit hyperbolic to denigrate the degree of consistency?

Thanks for slogging through this lengthy explication and responding to my