"polarized" terminology

[Pierre Deleporte]

Hi,

Dividing a character in two or more charater states is of course character 
coding, but more precisely it is character state delineation (the "bete 
noire" of phylogenetic inference according to Pogue and Mickevich 1990, and 
as underlined by Ken for morphological data).
Once you have delineated character states, you can "polarize the 
character", or more precisely polarize the transitions between its states. 
You can polarize characters (-states) a priori, before phylogenetic 
analysis, or a posteriori, as a result of the analysis: "checking" the 
optimal polarized evolutionary scenarios on the rooted cladogram.

In phylogenetic inference, ordering characters is a coding procedure 
consisting in giving a different weight to changes between different states 
of a character.   e.g. character A has three states 1,2,3. For the 
analysis, you can order it this way: weight 1 for changes between 1 and 2, 
and between 2 and 3, but weight 2 between 1 and 3. The implication is that 
the character has to pass through state 2 when changing between states 1 and 3.

It thus appears that you can order a character (weight differently 
different transitions between character states) without polarizing it (no a 
priori preference for the direction of evolution, e.g. from state 1 to 3 
versus from 3 to 1).

You can also combine a sophisticated ordering with polarization (e.g., 
stepmatrices of the program PAUP allow for weighting differently changes 
betwen character states according to the polarity of change: 2->3 may 
weight 1 while 3->2 weights 2).

best wishes
Pierre



At 14:16 06/01/2002 -0600, Ken Kinman wrote:
>    I'll have to think about the semantics of this one.  I was thinking that
>there is a difference between polarizing characters and polarizing
>"character states" (which is a special case of "ordering" character states).
>     What term would one apply to polarizing (dividing) a character into
>just two character states?  Whatever you call that kind of polarization, it
>needs to be done before one can perform a cladistic polarization (i.e., in
>the strict sense) of the "character states" (i.e. "ordering" sensu lato).
>Is Curtis referring to polarizing his characters, or is he just ordering the
>character states?
>     Don't mean to nitpick, but I've never really pondered the details of
>this terminology.  I'll really have to give this some more thought.
>         -------  Ken

>At 12:16 PM 1/6/02, Ken Kinman wrote:
>>     What term would one apply to polarizing (dividing) a character into
>>just two character states?
>
>Coding?
>
>
>--
>Curtis Clark                  http://www.csupomona.edu/~jcclark/
>Biological Sciences Department             Voice: (909) 869-4062
>California State Polytechnic University      FAX: (909) 869-4078
>Pomona CA 91768-4032  USA                  jcclark at csupomona.edu

Pierre Deleporte
CNRS UMR 6552 - Station Biologique de Paimpont
F-35380 Paimpont   FRANCE
Téléphone : 02 99 61 81 66
Télécopie : 02 99 61 81 88