Cladistics refers to methods of analysis for determining the relationships
among species.  It is based on the assumption that the important events are
the formation of two new species from an existing species.  This is not
necessarily how speciation happens (hybridization, splitting of multiple
species from the parent simultaneously, and a new species splitting off
from an existing one while the parent population shows no apparent change
are all exceptions), but is perhaps pne of the more frequent patterns.
        The analyses try to be more objective than the traditional approach
of deciding simply by the opinion of the person studying it.  Instead, the
data are organized into a list of features, called characters, and the
condition for these characters in each of the studies species, called
character states.  For example, color could be a character and red, blue,
yellow, and white could be character states.  Obviously, there is still a
subjective decision about what characters to include and how to code them
(for example, does a purple shell count as red or blue or both or a new
category).
        A mathematical analysis then tries to find what is the best
grouping of species to generate the evolutionary patterns observed.  These
calculations are computationally intense, and often approximations are
necessary.  The most popular method of analysis, maximum parsimony, tries
to arrange the taxa into a tree that has the fewest changes.  Following the
above example, if there were two blue species and two red species, it would
require fewer changes if the blue species were grouped together and the red
species were grouped together instead of pairing a blue and a red.  In real
data, there will almost certainly be conflicts between different
characters, so the analysis will end up suggesting that some character
states are convergently evolved.  Other approaches include distance
methods, in which a distance between species is calculated and the analysis
tries to find the shortest distance among the taxa, and maximum likelihood,
in which a model is developed specifying the probability that certain
evolutionary events will occur and then figuring out what arrangement of
species is most probable.  Each approach has its strengths and weaknesses
and rabid adherents, and it is still unclear how to test how good a given
set of results is.
        The more rabid cladists would like to throw out the Linnaean system
of nomenclature or modify it beyond recognition, because it was not
designed to convey cladistic information and in some cases can actually
obscure cladistic relationships.  However, the Linnaean system as currently
used does convey other useful information, and a diagram is much better
than a naming system for conveying cladistic information.
        My overall assesment is that cladistics are useful analytical
techniques as long as you keep sight of the basic goal of understanding the
evolution and relationships of organisms.
 
David Campbell
 
"Old Seashells"
 
Department of Geosciences
CB 3315 Mitchell Hall
University of North Carolina at Chapel Hill
Chapel Hill NC 27599-3315
USA
 
919-962-0685
FAX 919-966-4519
 
"He had discovered an unknown bivalve, forming a new genus"-E. A. Poe, The
Gold Bug