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


----- Original Message -----
From: <philidor11@snet.net>

> When I 'go out in the world and look at stuff', I'm apprehending
> reality.  The points we're both making are that reality exists
> and we can perceive it.  We've gotten past some major philosophical
> pettifogging.

While I'm at nitpicking respectively semantics... agreeing that there is a
reality is a _prerequisite_ for science and not part of science itself,

> How do you resolve two conflicting hypotheses?  You test them.
> How do you test conflicting phylogenetic hypotheses?  For starters,
> you
> combine the data in the two competing hypotheses and see what
> comes up.
> Perhaps it's one of the previous hypotheses, perhaps it's something
> new and
> different.
> [Ok.  What are the hallmarks of the test?  It is done with the
> same data (you said, '...combine the data),

He suggested that as _one possibility_, not the only way to do it. Other
possibilities? Run it through PAUP* and NONA and the tens of other programs.
Won't all have the same algorithms. Take the same organisms but use
different (or just more) characters. Add organisms. In the molecular world
use parsimony and neighbor-joining and minimum evolution and maximum
likelihood. Bootstrap the result and jackknife it and so on. To be precise,
the latter sentence is about testing the robustness of a tree -- how well it
represents the data it was made with, not about testing it against
additional data.

> and it uses the same
> logic as the initial analysis.

Not necessarily, see above.

> I would argue that what you are
> 'testing' here is how the results with a larger aggregation (sample
> of all relevant animals who ever lived) of data compare to the
> results from subsets of the data.  You'd certainly find out something
> about the computer program logic's response to expanding data,
> but the validity of the logic is untested.]

There are heaps of theoretical papers on that topic... unfortunately I
haven't read any. Would someone else like to jump in?

> The point I have been making, and will continue to make is that
> a
> phylogenetic hypothesis, a cladogram, is the DATA telling you
> what the hypothesis should and could be.  The just-so-story of
> Linnean
> systematicists is again, the anti-scientific way of looking at
> reality.
> Let's take the Linnaean view as one based purely on observation.

That's of course not true. Linné took an arbitrary part of the characters he
saw. "The just-so-story" refers to practically all pre-cladistic family
trees... HP Peter Buchholz was talking about methods _to find out
phylogeny_, not about ways to classify.

>  The definition of Aves = bird (shhh) is based on contemporary
> animals.  A cladistic definition of Aves includes the contemporary
> animals,


> a specific species (Archie) [or whatever else] included by assertion,


> and
> a mrca whose characters are subject to consensus change with
> peer-reviewed notice.

Why "consensus"? What characters the MRCA had is predicted by phylogenetic

> Now, a just-so story starts from the result and reasons back
> to the causes using comfortable premises.

It frequently does start with a few causes, draws a conclusion from these
that includes a whole evolutionary scenario, and then interprets all
remaining evidence so that it won't harm said conclusion.

> In other words, analysis
> forced to produce a pre-ordained result.

If you want to call suchlike an analysis.

>  I can assert that the cladistic definition is a just-so story
> because Archie MUST be included

When _something_ MUST be included, then the name has a stable meaning.
That's the sole purpose of phylogenetic definitions.

> and that it MUST be possible
> to identify the characters of the mrca.

I don't understand what you mean. Cladistic hypotheses make predictions
about some characters of every MRCA... let's find one, then they are tested

> I think the observational definition

What would that be? "I know it when I see it"?

> includes fewer assertions
> and arguable points, so therefore it is less of a just-so story.

Come on. :-)

> The birds you can see won't change, so a definition anchored
> on them is going to be unvarying.  That's the very opposite of
> subjective.

The birds won't change. But with time we'll see more and more organisms and
will have to decide whether to include them in Aves. When there's a fixed
phylogenetic definition all we need to do is to find out their positions in
the tree, then they either are in Aves or not. When we don't have one, we
can continue non-discussing our feelings like we did for centuries with few

> The problem is, the Objective Scientist Goggles (tm) have one
> type of polarization and the computer screen showing cladistic
> analysis has a glare shield with a different polarization.


> When
> our scientist searching for Truth looks at the cladistic screen
> through his Objective Scientist Goggles (tm), he sees complete
> darkness.  Someone, somewhere has made a decision during her/his
> cladistic analysis, and all chance of perfect objectivity has
> disappeared.

Nobody is perfect. Science is the best of all bad = subjective methods,
isn't it? :-)

> When the scientist, disappointed, looks away, and out the window,
> he sees a beautiful red cardinal near the window glass, and then
> sees and hears the bird's oranger mate screeching at him from
> a nearby branch for getting too close to the erratic humans.
>  The male cardinal stays on the window ledge; it's a cold day.
> Our scientist watches, enjoying.  He has found, if not THE answer,
> a perfectly good answer.

And then he asks himself what he has smoked, since InGen has finally managed
to clone *Coelophysis*, *Sinosauropteryx* and *Microraptor*, and he sees
them in his garden viciously biting at his "perfectly good answer", tearing
into 1000 little pieces.

> PS  Rereading the Jefferson quote lauding Linnaeus for having
> found a classification system everyone could agree upon,

Jefferson's predictive powers have turned out to have been limited.

> I found
> his statement regretting his error in momentarily agreeing that
> bones should be used to clarify distinctions.

Remember, Jefferson lived _a long time_ or 2 _ago_, _before Darwin had
published_, and thought it was indeed a totally subjective matter how to
classify organisms (and probably minerals -- remember Linné's 3rd kingdom is
Mineralia). Some artists, er, naturalists said using claws and horns for
systematics yielded the most beautiful result, others liked teeth better. It
was STAMP COLLECTING. Meanwhile it has been discovered that there is the
Tree of Life which links all organisms in only one pattern. So there is
actually one pattern that is (however marginally) more objective than all
others. When we make our classifications reflect the tree they become
testable. They become SCIENCE. Jefferson thought there was no scientific
possibility to classify -- too bad for him.

> Reminded me of John F. Kennedy's comment to a flock of Nobel
> laureates that there was more brain power in the White House
> that night than on any prior occasion, save when Jefferson dined
> there alone.
> Whatta guy!  (Jefferson)

Wasn't it he who said he'd rather believe that 2 Yankee professors lied than
that... whatever they said and has turned out to be correct? All due respect
to Jefferson, and that's much, but he wasn't a demi-god either.