ON TAXONOMIC MEMBERSHIP

Taxonomic membership frequencies exhibit distributions in which groups with fewer numbers of subtaxa are more common in a clade than those with more subtaxa. Here, a “broken-plate” model is developed to describe such taxonomic memberships in which some higher taxonomic group (the plate) is randomly subdivided into intermediate taxonomic units (plate fragments), whose sizes are dependent on the number of taxonomic subunits that each contain. Theoretical distributions of membership frequencies produced by this model yield a superior fit to data from both modern and fossil groups, as illustrated by classifications for primarily fossil brachiopods and entirely modern mammals. The nature of these distributions is consistent with the contention that Linnaean membership frequencies result from the random partitioning of taxonomic/morphologic space; the numbers of taxa contained within hierarchically equivalent groups are unrelated, as are membership numbers at taxonomically higher and lower levels of consideration.

Processes of evolutionary branching and/or practices of taxonomic allocation have, either jointly or independently, evidently served to give rise to schemes of classification in which clade members at equivalent taxonomic levels, as well as at successive hierarchical levels of consideration, are statistically unrelated. Agreement between observed taxonomic memberships and those anticipated from the random partitioning of diversity as described by the broken plate model bears directly on a number of fundamental questions including the significance of extreme polytypy and inferred adaptive radiation within many taxonomic groups. The notion that the numbers of subtaxa within any particular supertaxonomic group are largely indeterminate largely precludes their use as anything more than qualitative descriptors. While Linnaean classifications may imperfectly represent relationships among different clades, their taxonomic units remain the coin of the realm with respect to the ‘identification’ of living and especially fossil groups. Although the traditional Linnaean system has been particularly successful at categorizing diversity, it remains unclear whether these taxonomic groupings primarily reflect underlying natural patterns or superimposed anthropocentric preferences.