Clarifying the role of character loss in phylogenetic inference

Character loss, as opposed to absence, in the Annelida is regarded by Bleidorn (2007, Journal of Zoological Systematics and Evolutionary Research, 45, 299-307) as a problem for phylogenetic inference. Bleidorn's concern is derived from two misconceptions. First, from an epistemic standpoint, objects are perceived by way of their properties. 'Absence' refers to properties not observed of an object, rather than what is perceived, allowing an infinite possibility of positive instances. 'Loss' is an explanatory, not descriptive concept, directed at what is observed of objects. Neither absence nor loss is applicable to describing properties in need of explanation by phylogenetic hypotheses. The second problem is that partitioning data as 'morphological' and 'molecular' is not justified for discerning loss from absence. Naive partitioning is unwarranted since it violates the requirement of total evidence (RTE). The RTE precludes partitioning unless data are irrelevant to one another. Given that 'morphological' and 'molecular' data are used to infer 'phylogenetic' hypotheses, the relevance issue cannot be avoided, thereby denying partitioning. Ultimately, 'character loss' is not a problem once it is acknowledged that neither absence nor loss are properties of organisms that systematists perceive or explain by way of phylogenetic hypotheses.