PARSIMONY AND CHARACTER WEIGHTING

Abstract

Abstract— The likelihood justification of parsimony proposed in Sober (1983, 1984a) is applied to some problems posed by character weighting. An argument is provided for thinking that the point frequency of a character is not a good descriptor for parsimonious reconstructions of phylogeny. The idea that a good character will be conservative or nonadaptive is also examined from a likelihood point of view. © 1986 The Willi Hennig Society