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

Re: Cladists versus non-cladists

At 01:53 PM 5/8/97 -0600, Jonathan R. Wagner wrote:
>        What he means is, Evolutionary Taxonomy includes a great deal of
>information about "grade", similarity in morphology within a group. Thus, if
>A is more closely related to B than it is to C, it may still be grouped with
>C is it looks more like C than it does like B.

Now you are oversimplifying (which I admit I probably did with my
discussion of cladism and phylogenetic taxonomy - which is why I want some
help on my Web page).

The point is not to group them together merely because they *look* similar,
but to group them together if they share more fundamental biology with one
another.  Optimally, one would like to partition the evolutionary tree in
such a manner as to maximize the information content of the resulting
classification within the constraints of the linnaean hierarchy.  In this
way one can most reliably generalize biological knowledge about one member
of a taxon to another member.

> This may be due to
>convergence, reversal, etc. (collectively homoplaisy),

Nope.  That would make a polyphyletic taxon.  That is as verboten in
evolutionary taxonomy as in phylogenetic taxonomy.  Only phenetic taxonomy,
now mostly moribund, would even remotely consider allowing such a thing.

Only true shared characters, whether plesiomorphic or apomorphic, can be
used to characterize an evolutionary taxon.  Homoplaisy is right out as a
taxonomic character.

[Remember, I *do* advocate using cladograms as the first step in producing
a classification, even in evolutionary taxonomy - when Dr. Holtz's new
theropod cladogram comes out, I will probably use it to revise, or at least
check, my theropod classification on my Web page].

> or simply due to a
>lack of subjectively significant morphological similarity linking A 
>and B.

That is the old way of doing it.   Nowadays this is, or should be, done
using information theoretic measures on contiguous subsets of the nodes of
an evolutionary tree.

>ET workers judge this to be more useful than phylogenetic 

For *some* purposes, overall biological similarity *is* more useful.  For
others clade information is more useful.

>        By convenience of access, Stan means that, since in ET 
>animals are pidgeonholed into groups based on diagnostic characters, 
>one can easily and quickly find what group the animal belongs to. He
>also notes elsewhere that the hierarchical nature of a Linnean 
>structure, combined with the broad equivalence of (subjective)
>morphological difference in a Linnean tree makes it easier to rapidly 
>catagorize organisms. He notes that these features make
>ET a better information retrieval structure.

Pretty much.

I would, perhaps, emphasize identification and ecological and physiological
generalization a bit more.

I can (or could), with nearly 100% accuracy, identify a any North American
flowering plant to its family in about a minute in the field with NO
identification guide. (I am a little out of practice, so I may not be that
good anymore).

>        I should point out that the last sentence quoted above is a 
>bit unfair. While I can imagine Stan can quote some sort of figures 
>to say that cladists use more names, he seems to be proceeding from 
>the common misconception that every phylogenetic divergence must be 

Even allowing for that, the names are not clearly organized in a manner
that allows efficient identification and assignment.  In phylogenetic
taxonomy, as now practiced, one does not *have* a set of names of
equivalent rank such as the set of families of North American flowering
plants.  There is simply no mechanism for even determining that clade X and
clade Y are of "equivalent rank".  Indeed that phrase is essentially
meaningless in phylogenetic taxonomy.  Yet rank equivalence is one of the
perceptual organizational mechanisms that makes linnaean taxonomy so
useful. It allows focusing on a subset of the names most useful for a given
purpose (such as families for rapid field identification and rough
ecological estimation, or species for population studies and the like).

May the peace of God be with you.         sarima@ix.netcom.com