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

Re: Paleontology is science, not art

>>Although I understand you were never intending to
"belittle" paleontology, Toby, paleontology is not art.  It is a narrowly
defined scientific discipline aimed at posing answerable questions and then
finding ways, under the assumption of metholodical naturalism, to support or
falsify these questions.<<

Hi Matt!  I appreciated your nuanced discussion, but would you mind if I
suggest a couple more nuances?

First, describing an animal from a very small number of fossils is not a
simple mechanical process leading to a single, demonstrably best outcome.
To use your car mechanic analogy, there is no equivalent to seeing whether
the engine starts in paleontology.  If I were to study every way to
extrapolate clues from bones into a whole animal intensely enough to pass a
test and then analyzed a new  fossil, I would still prefer an experienced,
talented paleontologist's analysis of the fossil to my own.  The ability to
go from partial data to a whole solution is a rare talent.  The insight
required is an art.  Though Gary Kasparov was beaten once by a chess
computer, I'd still bet on him in a real match. In the same way, I don't
think anyone can program a computer, the type of  brute force analysis, to
match a top human in paleontology.
Given that definition of art as 'instinctual' insight, there is art in

>> These data might then be used in a cladistic analysis to yield a
probabilistic outcome of how closely these animals were related to other
known dinosaurs, which, of course, can always be falsified.<<

You can probably already see where I'm going with this one.  Extrapolation
of parts of a fossil to a whole can be confirmed by discovery of other
fossils from the same species, but there is an irreducible theoretical
element in cladistics.  Because the theory has to be persuasive, there is an
even larger role for finding the right 'feel' in a solution.
I've observed the discussions of character selections on the list closely,
watching knowledgeable judgement applied.  Sometimes people will observe
that the material being coded is a mess.  Sometimes they will observe that
the source fossils have been pictured vaguely or inappropriately described.
Sometimes they will write off a character's presence in two different
animals as a coincidence.  Sometimes they will say that a protuberance on
the fossil that appears to be one thing is actually something else.
Sometimes they will criticize another person's choice of animals to be
included in the analysis.  Sometimes they will note that there are a large
number of plausible outcomes; I think someone mentioned 750 in one case.
Sometimes they will wonder whether an animal should be so separate from all
the other animals they've found to be closely related, or wonder what it
means when an animal seems to be a reasonably good fit to two different
Sometimes they even observe that the selection and grouping of animals
included by someone else appears to be responsive to that scientist's
favorite theory of something or other.
I think that every one of the analysts producing cladograms is being honest
and dealing as well as they can with difficult data.  They are applying
their art.  I think too that some of these analysts will prove to be more
'right', in that their hypotheses will be more persuasive than others and
will serve as the basis for later work.  Those who produce 'better' answers
have arguably not followed the rules better than the others, but they have
applied a more discerning and compelling judgement.

This same difference in the quality of  theories obviously applies to other
sciences as well.  I'm not disagreeing with your basic point about the
scientific nature of paleontology of course, but I am suggesting a less
mechanistic phrasing about how analysis/examination of hypotheses is
conducted.  Any time you observe that a hypothesis *best* fits the data, I
see art and judgement being applied.
Thanks for another enjoyable discussion!