Defining the genus Homo

See allHide authors and affiliations

Science  28 Aug 2015:
Vol. 349, Issue 6251, pp. 931-932
DOI: 10.1126/science.aac6182

Almost 300 years ago, Linnaeus defined our genus Homo (and its species Homo sapiens) with the noncommittal words nosce te ipsum (know thyself) (1). Since then, fossil and molecular biology studies have provided insights into its evolution, yet the boundaries of both the species and the genus remain as fuzzy as ever, new fossils having been rather haphazardly assigned to species of Homo, with minimal attention to details of morphology.

What is “early Homo”?

(A to C, F, H) Specimens attributed to Homo with dental and mandibular features that resemble those of australopiths [e.g., posteriorly narrow, tapered molars; notch between cheek-facing/buccal cusps (*); anteriorly tall mandible], and (D, E, G) australopiths with features claimed to be specific to Homo [e.g., anteriorly narrow molars; no buccal-cusp notch; last molar not entirely masked by the ascending portion of the mandible (arrow)]. (A) OH7 (H. habilis); (B) KNM-ER 992 (H. ergaster); (C) Sangiran 6 (H. erectus); (D and G) Omo 75-1969-14A (australopith, Paranthropus aethiopicus), (D) occlusal, (G) lateral; (E) A.L.128-23 (A. afarensis); (F) Tighenif 1 (H. erectus); (H) Tabun II (H. neanderthalensis). The OH7, KNM-ER 992, Sangiran 6, and Tighenif 1 molars narrow posteriorly (A, B, C, and F), but in Omo 75-1969-14A and A.L.128-23 they are narrower anteriorly (D and E). A buccal notch (*) is present in KNM-ER 992 (B), Sangiran 6 (C), and Tighenif (F), but not in Omo 75-1969-14A (D) and A.L.128-23 (E). The last molar of Omo 75-1969-14A (G) would have been partially exposed in front of the ramus (arrow). The Tabun II mandible (H) is tall anteriorly. The fossils are not to scale.


In the late 18th century, Blumenbach attempted to distinguish H. sapiens—and by implication the genus Homo—from other mammals, citing morphological features such as small canines not separated from small incisors, a chin, a short mandible and face, a large rounded skull atop a vertically oriented spine, a bowl-shaped pelvis, and erect posture (2). Some of these features turned out to characterize only living humans, but others are found in all hominins. The 1856 discovery of the Feldhofer Grotte Neandertal provoked heated debate about human antiquity (3). At one extreme, Thomas Huxley saw the Neandertals as extending a continuum of living races into the past; at the other, William King argued that its distinctive morphology warranted a separate species, Homo neanderthalensis, for the first time expanding our genus beyond H. sapiens. Once Eugene Dubois had assigned human-like fossils from Java to Pithecanthropus erectus (now regarded as Homo erectus) in 1894, the human family rapidly accumulated new members, taking on the appearance of typically mammalian diversity.

But in the mid-20th century, the view of Homo changed dramatically. First, Dobzhansky (4) proclaimed that because culture allows hominins to adapt to any ecosystem, speciation in the human lineage was impossible. Then, Mayr (5) declared that as all known hominids were upright bipeds, all belonged to the single genus Homo. Further, his definition of Homo included only three Homo species—H. transvaalensis (the early hominids now known as australopiths), H. erectus, and H. sapiens—that were held to have transformed in sequence with time. The inclusion of the australopiths was rapidly rejected (6), but otherwise Mayr's extended concept of Homo, defined by adaptation rather than morphology, became rapidly entrenched in paleoanthropology (7).

In 1964, Leakey and colleagues (8) attributed the newly discovered ~1.8-million-year-old partial mandible, skullcap and hand (OH7) and foot (OH8), plus other materials from Olduvai Gorge, to the new species Homo habilis. This species replaced the very roughly contemporaneous South African australopiths in Mayr's transformationist scenario, although there was scant morphological justification for including any of this very ancient material in Homo. Indeed, the main motivation appears to have been Leakey's desire to identify this hominid as the maker of the simple stone tools found in the lower layers of the Gorge, following the dictum of Man the Toolmaker (9). This association has subsequently proven highly dubious (10). The inclusion in Homo of the H. habilis fossils so broadened the morphology of the genus that further hominids from other sites could be shoehorned into it almost without regard to their physical appearance. As a result, the largely unexamined definition of Homo became even murkier.

One strains to identify any derived features that would describe all or even most of this morass of specimens. Yet, subsequent attempts to define Homo have typically accepted the expanded genus (7, 11, 12). For example, Antón et al. recently united disparate East African, Georgian, and Asian fossils in H. erectus, prior to delineating two groups of “early Homo” (7). They excluded the H. habilis fossils from Olduvai from consideration because the OH7 mandible is distorted. One of their early Homo groups is exemplified by the Koobi Fora KNM-ER 1470 cranium and is derived in its tall, flat face and anteriorly straight-across maxilla that corners at the second premolar. The other, typified by the KNM-ER 1813 cranium, is distinguished by primitive (hence undiagnostic) features of its anteriorly rounded, narrow, and roughly parallel-sided maxilla. Had the authors not assigned the KNM-WT 15000 Nariokotome skull and skeleton to H. erectus (a species that is actually defined by younger, highly distinctive materials from Trinil and Sangiran, Java), an alternative might have been to note the striking facial, maxillary, and dental resemblances between KNM-ER 1813 and the Nariokotome individual. But either way, the ineluctable conclusion was that multiple species existed within “early Homo.

Shortly thereafter, Spoor et al. (11) virtually reconstructed the OH7 mandible, additionally extrapolating maxillary shape. As previously reported for the mandible (13), they found that the resulting jaws resembled those of australopiths and apes in being long, narrow, and roughly parallel-sided. Whereas Antón et al. had rejected brain size as a defining feature of their “early Homo” groups, Spoor et al. inferred from their reconstruction of the OH7 cranial vault that in early Homo, the jaws remained primitive, but the brain was enlarged. Noting dissimilarities both between H. habilis and other “early Homo,” and among the latter, Spoor et al. also argued that “early Homo” included multiple species.

Concurrently, Villmoare et al. (12) identified a ~2.8-million-year-old mandible with some teeth (LD350-1) from Ethiopia's Ledi-Geraru as the earliest member of Homo. Their reasons were that, unlike Australopithecus afarensis especially, its mandible is shallow anteriorly; its tooth rows are divergent; its ramus arises opposite the last, not second, molar; and its preserved M2 and M3 are narrower anteriorly, shallowly notched between the buccal cusps, and rounded posteriorly rather than tapering. They thus concluded that Homo had ancient roots and was diverse early on.

These studies clearly raise issues that will have to be clarified in any satisfactory morphological definition of the genus Homo and the species within it (see the figure). For example, mandibular shape is australopithlike in OH7 (panel A), KNM-ER 992 (panel B), and Lantian 1; the mandible is deep anteriorly in relatively recent Homo KRM1B/AP6222, Tabun II (panel H), and Qafzeh 9; the ramus arises opposite M2 in H. sapiens and such australopiths as SK55b and Omo 75-1969-14A (panels D and G); and, as in australopiths, the molars are notched buccally in KNM-ER 992 and H. erectus Sangiran 6 (panel C), while being broad anteriorly with tapered last molars in the temporally separated KNM-ER 992, Tighenif 1 and 2 (panel F), Qazeh 2, and KRM/AP6228-6230, all assigned to Homo. Conversely, in some undoubted australopiths, such as SK25, SKW5, and Omo 75-1969-14A, the preserved molars are LD350-1–like in being narrow anteriorly and/or flat buccally, as in Omo 75-1969-14A and Hadar A.L.128-23 (panel E) (13).

Recent analyses (7, 11, 12) agree in recognizing that the term “early Homo” masks taxonomic diversity. But they produce inconsistent results. Wood and Collard (14), for example, suggest that the species H. habilis and Homo rudolfensis be excluded from Homo for adaptive reasons, whereas Antón et al. (7) claim that an adaptive shift unites these species. Spoor et al. (11) stress the australopith-like shape of the H. habilis OH7 mandible, whereas Villmoare et al. (12) assert that the more derived features of LD350 warrant its inclusion in Homo. Nonetheless, all make clear how sorely both our concept of Homo, and the number of its known species, require reassessment.

The fact that detail has often been overshadowed by generalization in the study of the hominin fossil record is a product of paleoanthropology's unique history (3). But tradition alone cannot justify maintaining the status quo. Perhaps it is time to forget parochial habit and to begin approaching hominid systematics as students of other organisms do, especially in undertaking broader and morphologically more detailed comparisons than have been customary, and in revisiting morphological criteria for species recognition. If we want to be objective, we shall almost certainly have to scrap the iconic list of names in which hominin fossil specimens have historically been trapped, and start from the beginning by hypothesizing morphs, building testable theories of relatedness, and rethinking genera and species. In contrast to Mayr's austere linearity, we may find that human evolution rivaled that of other mammals in its evolutionary experimentation and luxuriant diversity.


View Abstract

Stay Connected to Science

Navigate This Article