ABSTRACT
Morphological evidence has been used to support the monophyly of both the Articulata and the Ecdysozoa. Although most recent computer-assisted cladistic analyses appear to support Ecdysozoa, several zoologists remain loyal to the classic Articulata concept. We address this phylogenetic debate from two perspectives. First, we discuss the striking differences in historical epistemology adopted by different workers, and how this inevitably leads to disagreement. Second, in order to provide a bridge of sorts between the different phylogenetic epistemologies we perform a set of morphological sensitivity experiments on various published morphological data sets to explore the robustness of the Articulata and Ecdysozoa hypotheses. We vary both the relative weight of characters, as well as the selection, coding, and scoring of characters. This approach allows a better insight into the relationship between character evidence and phylogenetic hypothesis for the different data sets. Depending on the data set, support for the Ecdysozoa varied from being weak or absent (Zrzavý et al. 1998; Nielsen 2001), somewhat stronger and moderately robust to changes in the data set (Zrzavý et al. 2001; Zrzavý 2003), to quite strong and robust to introduction of conflicting characters (Peterson & Eernisse 2001). However, by excluding problematic characters, correcting character coding and scoring errors, and introducing new potential articulatan synapomorphies the modified data set of Peterson & Eernisse (2001) yields support for a monophyletic Articulata. Ultimately, whether a given analysis supports Articulata or Ecdysozoa depends to a large degree on the phylogenetic philosophy that one adopts, and an unambiguous choice between these competing hypotheses, which will be accepted by all workers, seems therefore elusive.
