use of morphological characters

Pierre Deleporte Pierre.Deleporte at UNIV-RENNES1.FR
Wed May 3 16:19:05 CDT 2000

>At 08:39 02/05/00 -0400, Stefan Koenemann wrote :
> For example, my first character
>would be: appendage A = absent/present. 2nd character = whole appendage
>long/whole appendage short. 3rd character = reduced number of
segments/number >of segments not reduced. 4th character = proximal segments
narrow/proximal >segments dilated.
>In my opinion, this method is highly questionable. 
>First to note, all taxa with
>state 'absent' for the 1st character would have to be scored as '?' or '-'
>the following characters (2-4).

Yes, a "?" would be entered, meaning "inapplicable", and this is the best
thing to do, as advocated by Hawkins, Hughes and Scotland (1997) in
Cladistics 13 275-283 (however I suggest you consider separately the
"pattern cladistics" or "taxic approach" comments in this paper, which do
not bear directly on the problem).

>As far as I know, PAUP fills in the
>missing states with the most parsimonious solution, which would be
>contradictory in this case (an absent appendage cannot be short).

Platnick, Griswold and Coddington (1991) dealt extensively with this
problem of missing entries in Cladistics 7, 337-343. Undue "charging" of
missing entries with "impossible" states may lead to retain additional
minimum length trees which should be detected and discarded.

>Secondly, I would suspect the above characters, or at least some of them,
to >be dependent, e.g. the length of the appendage could be ontogenetically
>connected with a reduction of segments, etc.

Character independence is quite another and general problem. Lack of
character independence should hopefully be detected prior to the
constitution of the data matrix (on logic or biologic grounds), thus
avoiding redundant coding. But no doubt independence is more often assumed
than demonstrated (comments in Kluge 1989 "A concern for evidence..." Syst.
Zool. 38, 7-25).

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

More information about the Taxacom mailing list