A new matrix of postcranial hominin characters is here presented, consisting of 239 characters and 14 taxa. The taxa include extant humans, chimpanzees, and gorillas, and the fossil taxa consist of Ardipithecus ramidus, three species of Australopithecus, and seven Homo taxa. The matrix is 65.7% complete, with 2,197 scored character states. The characters are not polarized. These data could be useful for applications in cluster analysis, classification, and phylogeny.

A preliminary list of potential characters were derived from published descriptions of comparative data from the sources indicated in Table 1. The characters were then screened for redundant descriptions. Using published descriptions, scans, casts, and museum exhibits of original fossils, every character was assessed using relevant specimens. Where possible, missing character states were scored during this process. In some instances, the scans and photographs did not adequately replicate the original fossil, and character states could therefore not be observed. In those cases, character state assignments were left as originally reported and no new character states were recorded. For other characters, character descriptions could not be verified from examination of specimens, and new, related characters were devised in their place. Newly devised metric characters were assigned character states that roughly divided the range of observed values for all taxa in half. Finally, when character descriptions involved more than one bone, we split them into separate characters for individual bones. We introduced new characters based on descriptions and metric data reported by Prabhat et al. (2021), Prang (2016), and Lovejoy et al. (2016).

Table 1.  Initial sources of character descriptions.

Taxa were initially selected from a set used in previous fossil hominin studies (Dembo et al. 2015, 2016). Due to the dearth of available postcranial data, all Paranthropus species, H. rudolfensis, H. antecessor, Au. anamensis, Au. garhi, Sahelanthropus, and Kenyanthropus were not included in this character matrix. To maximize the number of character states assigned to each taxon, Asian Homo erectus and African Homo erectus (cf. Homo ergaster) were combined as a single taxon that we designate as Homo erectus sensu lato. Due to ongoing discussion about the diversity of fossils at the Dmanisi site and their affinity with H. erectus (e.g., Schwartz et al. 2014, Rightmire et al. 2019), we elected to treat the Dmanisi hominins as a single taxon “Homo georgicus” after Gabounia et al. (2002). Finally, we included characters obtained from published descriptions of the Little Foot skeleton (Clarke 2019a) in Australopithecus africanus sensu lato rather than a separate Australopithecus prometheus (see Berger and Hawks 2019, Clark 2019b, Grine 2019). None of these decisions should be considered taxonomic opinions but rather conveniences to maximize the number of known character states per taxon while retaining a good sample of taxa that represents the diversity of Australopithecus and Homo.

Here, the taxon sample consists of fourteen taxa: gorilla (Gorilla gorilla), chimpanzee (Pan troglodytes), modern humans (Homo sapiens), Neandertals, Homo heidelbergensis, Homo erectus s.l., Homo georgicus, Homo habilis, Homo floresiensis, Homo naledi, Australopithecus sediba, Australopithecus afarensis, Australopithecus africanus s.l., and Ardipithecus ramidus. Specimens of Homo heidelbergensis were largely derived from the remains from Sima de los Huesos, along with postcranial remains from Kabwe and the Berg Aukus femur. Homo erectus was based mostly on the Nariokotome skeleton along with remains from Indonesia. Homo habilis is based on individual specimens from Olduvai and sites in Kenya rather than any partial skeletal collection, and hence is the most likely to contain specimens that might be from other taxa (e.g. Homo ergaster/erectus or Homo rudolfensis). A fuller account of specimens and sources consulted is given in Appendix 1.

The full character descriptions are given in the accompanying Word file.

The characters are given in Nexus format, which is compatible with a variety of phylogenetic software.