Three-taxon analysis

[P.Hovenkamp]

At 12:53 AM 4/8/03 +0200, René Zaragüeta Bagils wrote:

(...)
>"Models do not enter, just evolution" :
>Lamarck's? Darwin's? Goldschmidt's? Neodarwinian's ? Bergson's? Teilhard de
>Chardin's? Vitalism? Orthogenetics? Budhist reincarnation?...
>If "just evolution" means anything, then "just evolution" is trivial.

There are some items in this list that I would not recognize as 
evolutionary - Vitalism, or reincarnation (and I don't really know 
Bergson). Apart from that, my point stands: if any of these proposed 
mechanisms results in a particular pattern of sister-group relationships, 
we may reasonably hope that we can find that pattern by studying the 
characters of the taxa. Of course, many of these mechanisms will result in 
very similar-looking patterns of relationship, and that is why there still 
may be controversy and uncertainty.

(...)
>- Second, you say that each character has had a state from which it
>developed.
>Then, the first one is a paraphyletic character (because it does not contain
>the state that developed from it). As you say, cladistic uses character
>analysis, which means that what we know about relationships of organisms is
>the same as what we know about relationships of characters. As you seem to
>defend paraphyletic characters (plesiomorphies, the states from which new
>characters
>develop) as informative, then they may only characterize paraphyletic
>groups, which do not exist in a Darwinian evolutionary world. I conclude
>that your evolutionary world is one where paraphyletic groups, characterized
>by paraphyletic characters, exist.
When I was at secondary school, we used to amuse ourselves by proving 
anything at all with series of impressive looking formula's, which on 
closer inspection always included a division by zero somewhere. All this 
elaborate reasoning strongly reminds me of that period. Once you start 
misapplying terms you can "prove" anything. Characters cannot be 
paraphyletic. The term does not apply. It is as simple as that.

>Example: Imagine that you tell me that tetrapod limbs developed from fins.
>This is the same that to say that tetrapods,
>characterized by limbs, arose from fishes, characterized by fins (instead
>of: tetrapods are a kind of fish characterized by limbs, a kind of fin).
>This
>is a general problem, applicable also to standard parsimony, of atomistic
>(transformations) versus epigenetic (differentiation) descriptions.
It's a problem only for those who fail to understand the relation between 
groups and characters. All others, when given the first statement (fins to 
limbs) would immediately conclude that any group defined only by the 
presence of fins is paraphyletic. In fact I think we agree on that - so 
what is your problem really?

>Concepts arise indeed de novo, in the mind of systematicians. Concepts are
>ideas. The concept of paraphyly arose de novo in the mind of Hennig, for
>instance, and has still to arise in lots of minds.
Debatable, but this should have a different thread. One might also say the 
the concept of paraphyly was already implicit in some earlier work, and 
that Hennig codified and named it. To trace back the history of the concept 
paraphyly would be an interesting subject for a historical study.

>Are you certain that characters cannot "arise de novo"? How do you know? How
>can I test your assertion?
Never ask for certainty unless you can offer it. You can test the assertion 
by looking for counterexamples. As you can any assertion.

>Perhaps you think that characters are observable bits of truth, but
>characters are ideas (hypothesis, theories), not observations.

And so is everything, until we become completely dizzy with the depth and 
profoundness of our views. Can we make any observations at all? If you say 
no, then retreat from science. If you say yes, you'd better start asking 
yourself how we do it instead of simply trying to deny we can.

> > Three taxon analysis gives up on character analysis even before it's
> > becoming difficult. Discarding evolution by the way.
> >
> > Peter Hovenkamp
> >
>Oops! You never practiced three-taxon analysis, didn't you?
Are you certain? How do you know? (I can play that game!)

>Perhaps one of
>the reasons of the lack of success of three-taxon analysis is its difficulty
>(conceptual and operational)!

Then solve those difficulties. You seem to accuse others of not doing the 
things that you cannot do yourself. Go ahead. Solve the conceptual 
difficulties of 3TA (some of them outlined above and in my other posts). 
Solve the operational problems. At least make a start. There is nothing in 
your post which addresses these problems.

Peter Hovenkamp