ABSTRACT
This chapter discusses the application of models originally developed for use in molecular sequences to the phylogenetic analysis of morphological data. The basic aspects of the molecular models are discussed first and then their morphological applications. The usual application of these explicit models is in maximum likelihood methods or Bayesian analysis. Both the basic methods to evaluate trees, as well as their potential problems and caveats, are discussed. All the methods used in standard practice assume a common mechanism—that is, some factor that jointly increases or decreases the probabilities of change along a given branch for all characters simultaneously. Empirical tests of this assumption in morphological datasets suggest that morphological characters do not evolve in such a manner. Under a wide array of realistic conditions, either relaxing or restricting some of the assumptions in the models, parsimony can be a maximum likelihood estimator, leading to preference for the same trees.
