The performance of morphological characters in broad‐scale phylogenetic analyses

In view of their propositional content (i.e. they can be right or wrong), character statements (i.e. statements that predicate characters of organisms) are treated as low‐level hypotheses. The thesis of the present study is that such character statements, as do more complex scientific theories, come with variable scope. The scope of a hypothesis, or theory, is the domain of discourse over which the hypothesis, or theory, ranges. A character statement is initially introduced within the context of a certain domain of discourse that is defined by the scale of the initial phylogenetic analysis. The doctrine of ‘total evidence’ requires the inclusion of previously introduced characters in subsequent studies. As a consequence, the initial scope of character statements is widened to the extent that the scale of subsequent analyses is broadened. Scope expansion for character statements may result in incomplete characters, in the subdivision of characters, or in ambiguity of reference (indeterminacy of the extension of anatomical terms). Character statements with a wide scope are desirable because they refer to characters with the potential to resolve deep nodes in phylogenetic analyses. Care must be taken to preserve referential unambiguity of anatomical terms if the originally restricted scope of a character statement is expanded to match a broad‐scale phylogenetic analysis. © 2007 The Linnean Society of London, Biological Journal of the Linnean Society , 2007, 92 , 297–308.