Classification standards such as the Mammal Species of the World (MSW) aim to unify name usages at the global scale, but may nevertheless experience significant levels of taxonomic change from one edition to the next. This circumstance challenges the
biodiversity and phylogenetic data communities to develop more granular identifiers to track taxonomic congruence and incongruence in ways that both humans and machines can process, i.e., to logically represent taxonomic provenance across multiple classification hierarchies. Here we show that reasoning over taxonomic provenance is feasible for two classifications of primates corresponding to the second and third MSW editions. Our approach entails three main components: (1) individuation of name usages as taxonomic concepts, (2) articulation of concepts via human-asserted Region Connection Calculus (RCC-5) relationships, and (3) the use of an Answer Set Programming toolkit to infer and visualize logically consistent alignments of these taxonomic input constraints. Our use case entails the Primates sec. Groves (1993; MSW2 - 317 taxonomic concepts; 233 at the species level) and Primates sec. Groves (2005; MSW3 - 483 taxonomic concepts; 376 at the species level). Using 402 concept-to-concept input articulations, the reasoning process yields a single, consistent alignment, and infers 153,111 Maximally Informative Relations that constitute a comprehensive provenance resolution map for every concept pair in the Primates sec. MSW2/MSW3. The entire alignment and various partitions facilitate quantitative analyses of name/meaning dissociation, revealing that approximately one in three paired name usages across treatments is not reliable - in the sense of the same name identifying congruent taxonomic meanings. We conclude with an optimistic outlook for logic-based provenance tools in next-generation biodiversity and phylogeny data platforms.
Classifications and phylogenetic inferences of organismal groups change in light of new insights. Over time these changes can result in an imperfect tracking of taxonomic perspectives through the re-/use of Code-compliant or informal names. To mitiga
te these limitations, we introduce a novel approach for aligning taxonomies through the interaction of human experts and logic reasoners. We explore the performance of this approach with the Perelleschus use case of Franz & Cardona-Duque (2013). The use case includes six taxonomies published from 1936 to 2013, 54 taxonomic concepts (i.e., circumscriptions of names individuated according to their respective source publications), and 75 expert-asserted Region Connection Calculus articulations (e.g., congruence, proper inclusion, overlap, or exclusion). An Open Source reasoning toolkit is used to analyze 13 paired Perelleschus taxonomy alignments under heterogeneous constraints and interpretations. The reasoning workflow optimizes the logical consistency and expressiveness of the input and infers the set of maximally informative relations among the entailed taxonomic concepts. The latter are then used to produce merge visualizations that represent all congruent and non-congruent taxonomic elements among the aligned input trees. In this small use case with 6-53 input concepts per alignment, the information gained through the reasoning process is on average one order of magnitude greater than in the input. The approach offers scalable solutions for tracking provenance among succeeding taxonomic perspectives that may have differential biases in naming conventions, phylogenetic resolution, ingroup and outgroup sampling, or ostensive (member-referencing) versus intensional (property-referencing) concepts and articulations.
