Marsupialia Illiger, 1811

Fox, Richard C. & Naylor, Bruce G., 2006, Stagodontid marsupials from the Late Cretaceous of Canada and their systematic and functional implications, Acta Palaeontologica Polonica 51 (1), pp. 13-36 : 15-16

Cohort Marsupialia Illiger, 1811

Remarks.—At present, little consensus exists as how best to classify marsupials (compare, e.g., Aplin and Archer 1987; Reig et al. 1987; Marshall et al. 1990; Szalay 1994; McKenna and Bell 1997; Kielan−Jaworowska et al. 2004; Case et al. 2004). Moreover, some recent authors ( Rougier et al. 1998; Luo et al. 2003; Horovitz and Sánchez−Villagra 2003; Asher et al. 2004) have preferred to limit the name “ Marsupialia ” to the crown clade, i.e., living marsupials, their last common ancestor, and all descendants of that last common ancestor. As a consequence, many major fossil groups of what for over a century have been considered marsupials by paleontologists and mammalogists have been relegated to a non−marsupial category informally termed “stem−Metatheria” or “basal Metatheria.” This revision is not owing to the anatomical characters that these “stem−” or “basal−metatherians” may or may not share with “crown marsupials”, but merely as a consequence of the crown clade definition being based on the occurrence of species in an arbitrarily selected time horizon, the Recent (Holocene). In actual practice, however, the scope of crown Marsupialia is sometimes even more limited than this, denoting only living species: contrary to the implications of crownclade definitions, none of the four papers cited above include in their analyses characters of extinct groups that descended from the “last common ancestor” of the crown clade. As an example of the consequences of such omissions, Luo et al. (2003) included loss of conules as an unambiguous synapomorphy of crown Marsupialia , while ignoring the well−documented presence of conules in Herpetotheriinae ( Korth 1994; Johanson 1996a, b). Herpetotheriines are Tertiary opossum−like marsupials that are near universally accepted as included in Didelphidae (e.g., Simpson 1945; Fox 1983; Marshall 1987; Reig et al. 1987; Marshall et al. 1990; Korth 1994; Johanson 1996b; McKenna and Bell 1997) and, hence, are crown clade marsupials even though extinct. The obvious fallacy here is that “commonness” in the living species was typologically assumed to be primitive for the crown clade and therefore must have characterized its last common ancestor and all descendants of that ancestor.

Regardless of the practices of individual paleontologists, however, the basic conceptual weakness of crown−group taxa is that they are defined by extinction events ( Lucas 1990; Miao 1991), an objection mostly ignored but valid nonetheless. By definition, crown−group taxa are those that have survived to the Recent, a criterion that muddles the distinction between adaptation of the organisms concerned and their descent relationships, yet it is the latter that furnishes the basis for classification in modern biology. Moreover, crown−group nomina are defined by reference to other nomina (definition by extension, i.e., a listing of items to which the definition applies), which in themselves have no material substance, leaving their reality impossible to demonstrate or refute. We believe that in order to be useful and subject to critical examination, membership in the units that are named, and hence the reality of those units and the utility of their definitions, can only be by reference to the material characters that they possess (definition by intension, i.e., by a list of properties required of all individuals included in the definition). In other words, for the working systematist, marsupials are marsupials because of the material features that allow their recognition and testify to their evolutionary history, not because of the taxonomic nomina that the name “ Marsupialia ” subsumes (see e.g., Kielan−Jaworowska et al. 2004). That being the case, the recent claim that no marsupials are known from the Cretaceous ( Rougier et al. 1998: 462) is based only on semantics, i.e., “language used to have a desired effect as in advertising or political propaganda” ( Mish 1983: 1068), not on the distribution of material characters that imply relationships among real organisms.