By G. Philip Rightmire [Department of Anthropology, Harvard University, Cambridge, MA 02138; and Department of Anthropology, Binghamton University (SUNY), Binghamton, NY 13902]

Approximately 700,000 years ago, Homo erectus in Africa was giving way to populations with larger brains accompanied by structural adjustments to the vault, cranial base, and face. Such early Middle Pleistocene hominins were not anatomically modern. Their skulls display strong supraorbital tori above projecting faces, flattened frontals, and less parietal expansion than is the case for Homo sapiens. Postcranial remains seem also to have archaic features. Subsequently, some groups evolved advanced skeletal morphology, and by ca. 200,000 years ago, individuals more similar to recent humans are present in the African record.

These fossils are associated with Middle Stone Age lithic assemblages and, in some cases, Acheulean tools. Crania from Herto in Ethiopia carry defleshing cutmarks and superficial scoring that may be indicative of mortuary practices. Despite these signs of behavioral innovation, neither the Herto hominins, nor others from Late Pleistocene sites such as Klasies River in southern Africa and Skhūl/Qafzeh in Israel, can be matched in living populations.

Skulls are quite robust, and it is only after ≈35,000 years ago that people with more gracile, fully modern morphology make their appearance. Not surprisingly, many questions concerning this evolutionary history have been raised. Attention has centered on systematics of the mid-Pleistocene hominins, their paleobiology, and the timing of dispersals that spread H. sapiens out of Africa and across the Old World. In this report, I discuss structural changes characterizing the skulls from different time periods, possible regional differences in morphology, and the bearing of this evidence on recognizing distinct species.

Stone artifacts and other traces of human activity dating from 700,000 to 130,000 years ago are found across Africa and Eurasia, and a number of sites contain well dated archaeological sequences. Fossils are far less plentiful, particularly compared with the abundant Late Pleistocene Neanderthals. Nevertheless, it is clear that some of the earliest populations differing from Homo erectus are documented at localities in Africa and Southwest Asia.

One important example is Bodo in the Middle Awash of Ethiopia, where a cranium, a broken parietal, and a humerus were discovered in conglomerates and sands containing mammalian fossils and later Acheulean tools. Radiometric dates point to an age of ca. 600,000 years (1). The cranium as reconstructed consists of the face and parts of the braincase.

There are resemblances to H. erectus in the massive facial skeleton, projecting brow, low and constricted frontal with midline keeling, and parietal angular torus. In other respects, Bodo is advanced in its morphology. Brain size is close to 1,250 cm3 (2) and substantially greater than expected for H. erectus. This difference is unlikely to result simply from larger body mass (3). Frontal squama proportions, the arched temporal contour, and some traits of the cranial base are like those of more modern humans. The browridge is divided into medial and lateral segments, the margin of the nose is vertical rather than forward sloping, and the incisive canal opens into the front of the palate. These are derived conditions present also in the face of recent Homo (4).

Another ancient cranium and a mandibular fragment were picked up at Elandsfontein in South Africa in 1953. Later, at the site designated Cutting 10, animal bones were uncovered with Acheulean bifaces, cores, and flakes. The fauna from Cutting 10 may not be associated directly with the artifacts, but the contemporaneity of many of the Elandsfontein bones with a later Acheulean industry is not in doubt (5).

The fauna includes bovids and other large herbivores, and there are archaic elements such as a dirk-toothed cat, a sivathere, a giant gelada baboon, and at least 4 archaic hartebeest/wildebeest-like antelope species. Some 15 of 48 mammalian species collected at Elandsfontein have no historic descendants, suggesting that this assemblage is 1.0 million to >600,000 years old (6). The human skullcap is cracked and weathered, but as with Bodo, resemblances to H. erectus are apparent. At the same time, the parietals are expanded (“bossed”), and the occipital is less angulated than in H. erectus. These differences are consistent with an increase in brain volume.

It is likely that the missing Elandsfontein facial parts are mirrored by the cranium from Broken Hill (Kabwe) in Zambia. The Broken Hill face is set forward from the anterior cranial fossa and exhibits very heavy brows. The cranial base is less flexed than is the norm for recent people. Despite the presence of these archaic features, the border of the nose is set vertically, and palatal anatomy is like that of later humans.

Also in its occipital proportions and in the temporomandibular joint region, Broken Hill shares derived traits with modern populations. Unfortunately, the cave deposits containing the fossils were long ago quarried away, and circumstances surrounding the 1921 discovery are no longer clear. Associations of the bones and artifacts are uncertain, but a tibia was found near the cranium. Efforts to date the hominins are underway (7), but for the moment the best indications are mammal fossils that suggest an age comparable to Bodo or Elandsfontein (8).

A Middle Pleistocene specimen from Salé in Morocco has proved enigmatic, because of distortion due to pathology (9). Another partially reconstructed braincase from Lake Eyasi in Tanzania is low in profile, although the upper scale of the occipital is vertical. The contour of the (left) parietal is rounded when the skull is viewed from the back. More of a cranium from Lake Ndutu is preserved. Reconstructive efforts (10) reveal a vault that is small (1,100 cm3) with side walls that are gently convex.

Also, the articular tubercle bounding the mandibular fossa is more prominent than would be the case for H. erectus, and the tympanic plate is delicate inferiorly, rather than thickened. The relatively gracile browridge, lack of strong mastoid cresting, and smooth nuchal region suggest identification of Ndutu as a female. A frontal bone from Zuttiyeh Cave in Israel has been interpreted both as an early Neanderthal and as a direct ancestor to the people at Skhūl and Qafzeh. Associated with Acheulo-Yabrudian artifacts ca. 350,000 to 300,000 years old (11), the Zuttiyeh fossil is more likely to represent an archaic population, similar to those in Africa.

Other fossils of later Middle Pleistocene age are linked with Middle Stone Age (MSA) tools. An example is Florisbad in South Africa, where bones and artifacts were recovered from a spring vent. The complex spring deposits and their contents have proved difficult to date, but ESR measurements on a human tooth give an age of ca. 260,000 years (12).

The Florisbad partial cranium, including the incomplete right side of a face, has been pieced together several times. Clarke's (13) reconstruction corrects some earlier errors, and it is evident that the face is more massive than had been supposed. The supraorbital torus is quite thick. However, the broad frontal is less constricted than that of Broken Hill. Hollowing of the maxillary wall (a “canine fossa”) has also been noted, but whether this morphology is “modern” remains uncertain, because the topography of the infraorbital surface is influenced by facial size (14). Unfortunately, neither upper facial height nor the extent of facial projection can be measured.