On the four complementary aspects of hierarchical character relationships and their bearing on scoring constraints, expressed in a new syntax for character dependencies.
Cladistics
; 39(5): 437-455, 2023 10.
Article
em En
| MEDLINE
| ID: mdl-37428134
Morphological matrices, including the conceptualization of characters and character states and scoring thereof, still are a valuable and necessary tool for phylogenetic analyses. Although they are often seen only as numerically simplified summaries of observations for the purpose of cladistic analyses, they also hold value as collections of ideas, concepts and the current state of knowledge, conveying various hypotheses on character state identity, homology and evolutionary transformations. A common and persistent issue in scoring and analysing morphological matrices is the phenomenon of inapplicable characters ("inapplicables"). Inapplicables result from the ontological dependency (based on hierarchical relationships) between characters. Traditionally handled the same as "missing data", inapplicables were shown to be problematic in holding the potential to result in unreasonable algorithmic preference for certain cladograms over others. Recently, though, this problem has been solved by approaching parsimony as a maximization of homology rather than a minimization of transformational steps. We herein aim to further improve our theoretical understanding of the underlying hierarchical nature of morphological characters, which causes the phenomenon of ontological dependencies and, thereby, inapplicables. As a result, we present a discussion of various character-dependency scenarios and a new concept of hierarchical character relationships as being composed of four complementary sub-aspects. Building on this, a new syntax for the designation of character dependencies as part of the character statement is proposed, to help identify and apply scoring constraints for manual and automated scoring of morphological character matrices and their cladistic analysis.
Texto completo:
1
Base de dados:
MEDLINE
Assunto principal:
Conhecimento
/
Evolução Biológica
Tipo de estudo:
Prognostic_studies
Idioma:
En
Revista:
Cladistics
Ano de publicação:
2023
Tipo de documento:
Article
País de afiliação:
Alemanha