Re: Underlying basis of classification (Was: Re Dinobirds)

[chris brochu <cbrochu@fmppr.fmnh.org>]

Philidor11@aol.com wrote:
> 
> In a message dated 7/19/99 9:11:55 AM Eastern Daylight Time,
> cbrochu@fmppr.fmnh.org writes:
> 
> <<And the computer algorithm actually creates an unrooted network before we
> specify an ancestor - we really can't interpret ancestors on an unrooted
> network, since we don't know which way time runs on something like that.>>
> and
> << The time axis is secondary to the network and imparts a sort of
> "directionality" that we can interpret as the result of evolution. >>
> 
> I see it.
> The only way an ancestor could not be created is if the diagram worked
> without descent, providing a circular constellation of animals based on
> certain specified traits.

Uh.....no.  In a four-taxon unrooted statement, two taxa may be
"joined," even though they are not sister taxa and it would be wrong to
infer their branching point as a common ancestor.  One could even root
it in the middle of the internal branch.

Unrooted networks have branching points - only we don't know which ones
are going to be nodes joining sister taxa in the rooted tree.

Have a look at http://www.sinauer.com/Titles/Text/judd.html#chaps
(select chapter 2), especially figures 2.4 and 2.6, for examples of
unrooted networks.  These are, unfortunately, very complicated diagrams
- the phylogeny reconstruction chapter in Hillis et al.'s Molecular
Systematics (Sinauer Press) discusses simpler diagrams.



  The model of basal/derived traits implying closer
> to/farther from an ancestor is a separate, secondary analysis.  Presumably,
> if this model could not be successfully applied, then the observed
> association of the animals would be rejected. 

The word "model" here is problematic, since it's more commonly used in
likelihood analyses rather than parsimony and implies prior knowledge of
evolutionary process.  "Distribution of character states on the tree"
would be slightly more appropriate.




 Given the possibility of
> regression to more 'primitive' characters and the fact that the characters do
> not have to be combined for a functional purpose in a single animal, this
> secondary analysis is probably very forgiving.
> My difficulty in seeing this was the assertion in previous discussions that
> cladistics is a classification system based on evolution in preference to
> description.


In some ways, it is.  "Cladistics" is a word that can be used in a
variety of ways - for a parsimony analysis (how I use it), for
phylogenetic taxonomy (what you imply above), for numerical
phylogenetics (this would make a great many people bristle), etc.  Since
we morphologists use parsimony to the near-exclusion of anything else,
confusion of cladistics with numerical phylogenetics is understandable.

What we've been discussing in this thread is phylogeny reconstruction -
the parsimony analysis version of "cladistics."  Taxonomy is a secondary
issue to this.



  From this it seemed to follow that any connection among animals
> must first be consistent with an evolutionary progression.  Methodologically,
> this is simply not true.  Shows you have to check your premises carefully in
> following an argument.
> 
> By the way, I  want to understand your point on the group:
> 
> <<(Amoeba(Oak Tree(Starfish, Tyrannosaurus)))
> Where are the ancestors here?  Is the comma between the starfish and
> teerex an ancestor?  The parenthesis between the oak and the metazoans?
> One could also express them with internested circles or boxes.>>
> 
> Leaving aside my initial feeling that you were saying that the oak tree was
> ancestral to tyrannosaurus, a relationship complicating family reunions, and
> that a single symbol can cover multifarious gangs of ancestors, I wonder if
> you want me to read the punctuation as 'the oak tree split off from a line
> leading to starfish before the line leading from starfish to Tyrannosaurus.'

Sort of - though a phylogeneticist would more likely read it as "the
starfish and teerex are closer relatives to each other than either is to
the oak tree."


chris


-- 
----------------------
Christopher A. Brochu
Department of Geology
Field Museum of Natural History
Roosevelt Road at Lake Shore Drive
Chicago, IL 60605

voice: 312-665-7633  (NEW)
fax: 312-665-7641 (NEW)
electronic:  cbrochu@fmppr.fmnh.org