An ontological dilemma: Epistemology and methodology of historical biogeography

The task of historical biogeography is to reveal and explain the history of biotas and their historical connections. Historical 'relationship' between biotas is defined as the sharing of descendants of the same ancestor. Several generalized patterns of relationship, rather than a single universal one, should be expected for any set of biotas or areas. Such generalized patterns must be sought by comparison of individual patterns, based on individual monophyletic groups. Generalized patterns of biota relationships are formally statements about numerical universals, while a pattern derived from an individual group is a statement about a particular. Such statements are not subject to testing in the Popperian sense; they may be falsified as well as verified. As in other historical sciences, explanation in historical biogeography is genetic in form and probabilistic in nature. An explanation is falsified when an explanatory premise is, and corroborated when an explanatory premise is verified. Two major methodologies exist to derive 'area cladograms' from substituted (taxon) cladograms, Component Analysis and Parsimony Analysis. It is argued that Parsimony Analysis is the theoretically more satisfactory of these because it treats all sources of 'error' in the same way, making no process-related assumptions about the sources of conflicts. The rooting of area dendrograms is problematic. The rooting by an 'allzero' outarea is theoretically unsound and the dendrogram should be rooted a posteriori adjacent to the ancestral area(s), or not at all.