>By choosing an appropriate set of criteria
>which is applied to *all* the species in a genus, the entire concept of
>a key is instantly changed. You see, suppose I wished to key out the
>non-anthocyanin-free _Sarracenia_ (recognizing no taxa beyond species).
>And suppose my criteria was as follows (using 1=yes)
>(...) normal botanical keys have the most important distinctions keyed out
>>first (...) There are strengths and weaknesses to this method, I guess.
What you suggest is a table rather than a key. This is indeed a method
related to cladistics. The drawback of cladistics is the fact that *no*
weighting of the characters can be distilled from the table (the weighting
has rather to be a part of the input). The advantage is of course some sort
of transparence for your given phylogenetic reasoning (and the creation of
phylogenetic trees can be performed by a computer).
I prefer (like Peter G. Taylor, S. Jost Casper, Elza Fromm-Trinta, Ludwig
Diels, John M. Macfarlane, and Benedictus Hubertus Danser) the
nineteenth-century method of studying the plants rather than studying
character tables. (expert- based system vs. expert-system base). The
drawback with this method is the fact that you sometimes cannot give
perfect answers to the question how you arrived at your results (and you
cannot have the job of taxonomy done by a computer).
I have tried cladistics with _Nepenthes_, once. But the problem was proper
selection of suitable characters and character states, and of course
computing time. My impression is that the output is perfectly apt to
confirm prejudice; e.g. my favourite hypothesis concerning _Nepenthes_ is
the _N.maxima_ group (species with a conspicuous projection on the lower
surface of the lid near the base, but excluding _N.alata_) as a natural
entity. Under no circumstance did the program distribute these species
other than within their cluster. But this was certainly caused by the data
matrix used. The program just doesn't see more than the person who did the
input. Anyway, the program did not produce an unambiguous tree, and no key
did emerge from that attempt.
Kind regards
Jan