Posted by Mike Dunford on July 16, 2005 07:43 PM

Today’s entry comes from the “creatively nasty” file:

Lorentzen and Sieg (1991) have insisted that parsimony analysis should properly result in the display of all possible assignments of ancestral nodes (and thus all possible placements of changes of state) for each character. Their objective may find little support outside of the pulp and paper industry.
  —Joeseph Felsenstein
  Inferring Phylogenies (p70)