Homo ergaster and Its Contemporaries … · systematics and taxonomy of the genus Homo in the Early...

Homo ergaster and Its Contemporaries

Ian Tattersall*

Division of Anthropology, American Museum of Natural History, New York, NY, USA

Abstract

On the basis of their strong morphological differences from the Javan type materials, manyauthorities now consider the diverse East African fossils initially classified as “African Homoerectus” to be more properly allocable to the speciesH. ergaster. However, while this separation atthe species level of the African and Indonesian hominids is certainly justified, the species H.ergaster as thus constituted still embraces a significant morphological variety. Indeed, althoughthis grouping of African fossils seems to form a fairly coherent clade, it also appears quite diverse.The East Turkana type mandible of H. ergaster is matched by other specimens from Kenya andTanzania, but not by the mandible of the iconic WT 15000 skeleton, and in its turn this specimenfails to match either in its cranial construction or its upper dentition most of the other comparablespecimens usually referred toH. ergaster. Clearly there is a need for a systematic reappraisal of theentire “African Homo erectus” ¼ Homo ergaster group, and equally clearly the hominid evolu-tionary story throughout the Old World in the Early–Middle Pleistocene was more complex than isimplied by the extension of the species H. erectus to cover the entire miscellaneous assemblage ofhominid fossils from this time period.

Introduction

There is probably no area of paleoanthropology in which disagreement is more profound than in thesystematics and taxonomy of the genus Homo in the Early to Middle Pleistocene. This discord hasa long and, dare one say it, illustrious pedigree, dating right back to the initial discovery anddescription of the Javan species Pithecanthropus (¼Homo) erectus by Eugene Dubois in the early1890s. At that time the only extinct hominid known was the European H. neanderthalensis, a formthat, though peculiar in morphology, possessed a brain of modern human size. The new and moreancient hominid announced by Dubois as an intermediate between modern humans and apes (astatus reflected in his initial choice of name, which translates as “upright ape-man”) was thus thefirst known human fossil relative to display a brain cavity that was significantly larger than those ofmodern-day great apes while lying below the range ofH. sapiens. Dubois’s discovery unleashed animmediate furor. The key to Dubois’s interpretation of this specimen as a human relative (thoughhe stopped short of placing his find in the human family Hominidae) was the association of the typeTrinil skullcap with a femur whose morphology was without doubt that of an upright biped in themodern fashion. This association was immediately attacked (and has continued to be periodically

*Email: [email protected]

Handbook of Paleoanthropology

DOI 10.1007/978-3-642-27800-6_53-5# Springer-Verlag Berlin Heidelberg 2013

of 18

questioned), initially by those who preferred to see the cranium as that of a specialized ape, mayberelated to the gibbons. At the same time, many of those who accepted the association between thecranium and femur wrote off the former as deriving from an aberrant modern human (seediscussion in Tattersall 1995, 2008).

Still, some paleoanthropologists (Cunningham 1895) did seize immediately upon the Trinilspecimen as an evolutionary intermediate between great apes and humans, and were willing toview the Javan hominid as an early member of a lineage that had given rise to H. sapiens via theNeanderthals. This interpretation rapidly gained ground (Theunissen 1988), and by early in thetwentieth century, not least through the efforts of Schwalbe (1899) – and despite those of Boule(1911–1913) – the place ofH. erectus as the “hominid in the middle” had effectively been secured.Given the tiny size of the hominid fossil record at that time, and that the apparent rudiments ofa transformation series in brain size were present in what was known, this interpretation was hardlysurprising: indeed, it was a good story that was hardly contradicted by the few facts then available.And, in the decades before the Second World War, two additional developments conspired to keepH. erectus at the front and center in scenarios of human evolution.

The first of these was the discovery of the huge trove of Sinanthropus pekinensis fossils atZhoukoudian near Beijing during the late 1920s and the 1930s, and of the similarly impressiveseries ofH. soloensis crania at Ngandong in Java in 1931–1932. The hominids from both sites werereckoned to be very close to JavanH. erectus, if not exactly the same thing; and at a time whenmosthominid fossil sites produced a specimen here and there, both discoveries were overwhelming byvirtue of the sheer volume of material produced. At the same time geneticists, systematists, andpaleontologists in the USA and Europe were busily constructing the outlines of what came to beknown as the Evolutionary Synthesis, which saw the gradual modification of continuous lineagesas the central feature of the evolutionary process (see discussion in Tattersall 1995). And at mid-century, the ornithologist Ernst Mayr (1950), one of the principal architects of the Synthesis,bluntly told the paleoanthropological profession that S. pekinensis, H. soloensis, and other MiddlePleistocene hominids all belonged to H. erectus, the species that occupied the middle part ofa direct and gradually transforming lineage running from H. transvaalensis (the australopiths) atthe beginning to H. sapiens (which embraced the Neanderthals) at the summit.

Mayr’s short article was perhaps the most influential contribution ever in paleoanthropology,and effectively set its agenda for the next half-century. The rapidly increasing size of the humanfossil record eventually forced even Mayr to relent, and to admit a little more complexity into thepicture; but for decades, paleoanthropologists labored steadfastly under the notion that the evolu-tionary history of our kind had largely involved the gradual modification through time of a centrallineage that eventually culminated inH. sapiens. Of course, it was admitted that at any one point intime such a lineage, widely distributed across the Old World, would have harbored a variety oflocal variants (see chapter on “▶DefiningHomo erectus” by Baab); but throughout the second halfof the twentieth century, the emphasis was principally on within-species variation, rather than onthe question of whether a signal of systematic (species) diversity might be detectable in the varietyof morphologies that emerged as the hominid fossil record steadily enlarged. Against this back-ground, the category H. erectus became a catchall for a huge and unwieldy assortment of fossils ofsubstantially differing morphologies.

Such hominids came fromwidely scattered localities. First, the probably 1.0- to 0.7-million-year(myr)-old Trinil specimens from Java were joined by a steady stream of discoveries in the nearbySangiran Dome, not far away, that probably date in the 1.5- to 1.0-myr range, most of them closer toits younger end. Then the sample was augmented by the Chinese Peking Man fossils, now thoughtto be probably between about 500 and 300,000 years (kyr) old, followed by the Ngandong



of 18

http://dx.doi.org/SpringerLink:ChapterTarget

specimens (which may be as young as 50–30 kyr old), and ultimately by other Javanese fossils fromlocalities such as Sambungmacan and Ngawi, both uncertainly dated but unlikely to be more than200 kyr old, and most probably younger. In China, later finds attributed to H. erectus came fromsites including Lantian (Gongwangling and Chenjiawo, both perhaps around 1.0 myr), Hexian(maybe 400 kyr), Nanjing (ca. 350 kyr), and even Longgupo, a site that may possibly be as much as1.8 myr old. Some European specimens in the 400–300 kyr range, such as those fromVertesszollos(Hungary), Arago (France), and Bilzingsleben (Germany), have been referred by some authors toH. erectus, as has the 900- to 800-kyr-old calvaria from Ceprano in Italy. Further east, in theCaucasus, the 1.8-myr-old Georgian site of Dmanisi has yielded fossils that have also beenattributed to H. erectus. In Africa practically anything from the earlier Middle Pleistocene, andsoon many older specimens as well, found themselves identified as H. erectus, so that the speciescame to include such motley fossils as the 1.4-myr-old Olduvai Hominid (OH) 9 calvaria fromTanzania; the 700-kyr-old mandibles from Tighenif (Ternifine) in Algeria; the 400-kyr-old Salepartial braincase from Morocco; the 1.6-myr-old WT 15000 “Turkana Boy” skeleton fromNariokotome on the western side of Lake Turkana in Kenya, and several crania and mandibles inthe 1.9–1.5-myr range from Koobi Fora and Ileret on the eastern side of the same lake; also fromKenya, the fragmentary Olorgesailie hominid at 1.0–0.9 myr; the 1-myr-old Daka calvaria fromEthiopia; a cranium of apparently similar age from Buia in Eritrea; and even, from South Africa,the perhaps 1.6-myr-old Swartkrans Member 1 SK 847 partial cranium.

Not only do these fossils cover an enormous span of time (ca. 1.8–0.03 myr), but they alsoembrace a huge range of morphologies, and taken together they hardly suggest a neat chronologicalseries of the kind the Synthesis had predicted. Clearly, there is a systematic signal of some kind inthe assemblage of hominid fossils that have at one time or another been allocated toHomo erectus:a signal of diversity at the species as well as the morphological level. But it was not until thebeginning of the final quarter of the twentieth century that this possibility began to be seriouslyinvestigated.

Enter Homo ergaster

The existence ofH. erectus as a convenient catchall for a remarkable variety of hominids certainlyfacilitated the telling of a relatively simple and straightforward human evolutionary story that couldbe told in terms of consistent long-term selection pressures for such things as more perfectthermoregulation, more efficient digestion, and above all greater intelligence. However, thisstory of the gradual honing over time of an ever more effective human machine was contradictedby the growing post-Synthesis realization that the evolutionary process consists of a great dealmore than simple natural selection (Tattersall 1995, 1998). It also sat rather uneasily with the factthat the Pleistocene was increasingly being seen as a period of extraordinary short-term climaticoscillations as well as of longer-term fluctuations (see chapter on “▶Quaternary Geology andPaleoenvironments” by Van Couvering). And, most significantly of all, it was not the story that thegrowing assemblage of Homo fossils seemed to be telling.

The first shot across the bows of the all-encompassing notion of H. erectus came from work onfossils found at the classic African localities on the eastern shores of Lake Turkana. During the1970s, several fossils, notably the partial crania KNM-ER 3733 and KNM-ER 3883, werediscovered at Koobi Fora that their describers (Leakey and Walker 1976; Walker and Leakey1978) ascribed to the species H. erectus. With cranial volumes of 848 and 804 ml respectively(these andmost of the other endocranial volumes cited here come fromHolloway et al. 2004), these



of 18



1.8- and 1.6-myr-old individuals had possessed brains almost as large as that of Dubois’s muchyounger (1.0–0.7 myr) Trinil type specimen of H. erectus, at about 950 ml. Another significantspecimen was a well-preserved mandible, KNM-ER 992 (Fig. 1), about 1.5 myr old, which wasrecovered at Ileret and described simply as Homo of indeterminate species (Leakey 1972). Soonthereafter, Groves and Mazak (1975) jumped into the fray and made ER 992 the type specimen ofa new species,H. ergaster. Although this innovation was disdainfully dismissed by the Koobi Forateam, and some other influential workers (Rightmire 1990) also continued to prefer the morecomprehensive concept of H. erectus, this move finally opened the door to a reappraisal of thetradition of automatically assigning to H. erectus any and all African fossils with measured orassumed brain sizes in the general range of those noted above. In short, it became possible toentertain the notion that H. erectus is a terminal eastern Asian hominid species, and that hominidevolution throughout the Early andMiddle Pleistocene continued in the pattern already established,with a vigorous exploration of the many different ways in which it was possible to be a hominid inthe shifting and highly varied habitats of the Ice Age Old World. The next section, on “CranialMorphologies in Homo of the Early to Middle Pleistocene,” will examine the morphologicalevidence for diversity within the immediate group to which bothH. ergaster andH. erectus belong.

Cranial Morphologies in Homo of the Early to Middle Pleistocene

General ConsiderationsNobody doubts that the H. erectus/H. ergaster group represents a relatively cohesive subset of thefamily Hominidae. The question, clearly, is whether in this group of fossils we are looking ata radiation of species or at a single hugely variable species that may or may not have evolveddirectionally over the entire expanse of time (about 1.9–1.8 to 0.03 myr) and space (virtually theentire habitable Old World) it occupied. The distinction here is an important one, for species (evenif not greatly differentiated morphologically from their closest relatives) are historically individ-uated entities which can compete with one another for ecological space and become extinct withouttrace, whereas within species even demes that are significantly differentiated morphologicallyremain ephemeral entities that can disappear simply by absorption into ongoing conspecific

Fig. 1 Lateral view of the KNM-ER 992 holotype mandible of H. ergaster. Scan of cast by Ken Mowbray



of 18

populations. The twin processes of speciation and genetic/morphological differentiation are notlinked, so that speciation may take place in the absence of significant morphological divergence,while the latter can occur without speciation intervening. This, of course, often makes unequivocalspecies recognition difficult in fossil assemblages (Tattersall 1986). However, it seems generally tobe the case among living primates that, where substantial osteological differences are presentamong populations, those populations tend to act in nature as distinct species (i.e., as effectivelyindependent reproductive entities). If we apply this criterion very conservatively to speciesrecognition in the fossil record, demanding that the fossil species we recognize consistently beardistinctive osteological differences from related forms, we will probably underestimate the numberof species in that record. However, we will not distort its overall pattern (Tattersall 1986, 1992). Itis important to bear in mind that not all “morphs” (distinctive morphological entities) that werecognize in the fossil record will necessarily correspond to species in the reproductive sense; but itis equally evident that, given the nature of the fossil record, morphology must be the starting pointin our analyses of it. After all, neither geological age nor geographical provenance, the other twoattributes of any fossil, is necessarily linked to species identity, while its morphology is the onlyfeature that makes a fossil species recognizable at all. It is in this spirit that the remainder of thissurvey is offered, with the proviso that we will clearly not learn much that is useful about thepattern of events in hominid evolution during the Pleistocene if we do nothing more than replacethe term “African Homo erectus” with the equally sweeping H. ergaster.

Eastern AsiaGiven the established conventions of nomenclature and systematics, when we begin to consider themass of material that has at one time or another been allocated to H. erectus/H. ergaster we mustnecessarily, begin with Dubois’s holotype material (Fig. 2) from Trinil in Java (Schwartz andTattersall 2003, 2005). It is the morphology of the Trinil skullcap that defines the species H.erectus, and the allocation of other fossils to this named entity must be done on the basis of theirmorphological similarities to it. The problem lies, of course, in deciding just how close thosesimilarities should be, and it has to be admitted that there is no quantifiable answer to this question.

The Trinil 2 holotype is highly derived among hominids in a number of characteristics,especially of the brow region and the rear of the skull (Schwartz and Tattersall 2000, 2005). It isa smallish, long, thin-boned calotte with a narrow, shelflike, and laterally flaring postorbital regionthat flows onto the long, gently sloping, and flattish frontal plane with an almost imperceptiblemidline keel defined by shallow depressions bilaterally. The lateral walls of the low braincase are

Fig. 2 Lateral view of the Trinil holotype calotte of H. erectus (Photo courtesy of #Jeffrey H. Schwartz)



of 18

short and are markedly tilted inwardly above faint, low-set temporal lines; and the rather acutenuchal angle is distended posteriorly into a well-defined horizontal torus. Although the hominidsample from the adjacent Sangiran region is quite variable, especially in robusticity, the basicTrinil braincase morphology is repeated in most of the crania, the major exception being the quitecomplete if somewhat distorted cranium Sangiran 17 (Fig. 3). The Sangiran sample of upper and

Fig. 3 Lateral views of crania from Africa and Eurasia in the “Homo erectus/Homo ergaster” group. Left column, topto bottom: Sangiran S17; Koobi Fora KNM-ER 3733; Koobi Fora KNM-ER 3732; West Turkana KNM-WT 15000;

Dmanisi D 2282. Right column, top to bottom: Zhoukoudian cranial reconstruction (Sawyer and Tattersall version);

Koobi Fora KNM-ER 3883; Olduvai OH 9; Koobi Fora KNM-ER 1813 (scale ¼ 1 cm) (From Schwartz and Tattersall

(2005). Photo courtesy of #Jeffrey H. Schwartz; Turkana fossil images courtesy of National Museums of Kenya)



of 18

lower dentitions is also heterogeneous, suggesting that a second hominid morph may be present inaddition to the Trinil one.

The two (now three) crania known from the Sambungmacan region, to the northeast of Sangiran,are substantially younger than the Trinil/Sangiran assemblage and share a contrasting browstructure in which the quite horizontal supraorbital tori thicken laterally and appear to be contin-uous across glabella. The braincase itself has the appearance of being rather better inflated thantypical of the Trinil form (the two published brain volumes are 1,035 and 917 ml), and the coronalprofile is tent-shaped rather than having a squat and rounded outline. The nuchal plane undercutsthe occiput to produce a horizontal torus that is well defined below, but has much poorer definitionabove. Together these specimens, along with another calvaria from Ngawi (870 ml), producea morph that generally resembles the Trinil type but is readily distinguishable from it. TheNgandong crania are similarly distinctive. Larger, and more robust than the others, withendocranial volumes that range from 1,015 to 1,250 ml, they differ from the Trinil form in theways in which the Sambungmacan ones do. However, in addition they present yet more capaciousbraincases that have more or less vertical side walls with quite aggressively raised temporal lines;and they show a greater rearward projection than the Sambungmacan forms do of the occipital

Fig. 4 Occlusal views of mandibles. Top left: left side of Koobi Fora KNM-ER 992. Top right: Koobi Fora KNM-ER

3734. Middle left: Olduvai OH 22; Middle right: left side of West Turkana KNM-WT 15000. Bottom: left side of

Dmanisi D 211 (scale ¼ 1 cm) (From Schwartz and Tattersall (2005). Photo courtesy of #Jeffrey H. Schwartz;

Turkana fossil images courtesy of National Museums of Kenya)



of 18

torus, a feature with a well-defined superior border. Nonetheless, this entire group is united, inparticular, by a set of derived supraorbital and nuchal morphologies, and it presents itself asa relatively cohesive whole. Clearly, these Javan forms are part of the same eastern Asian hominidclade and, if it is not divided up (basically, at this point, a matter of taste), it is this assemblage thatmust provide the core identity of the species H. erectus.

The hominid fossils from Locality 1 of Zhoukoudian, in China (Fig. 3), have generally beenconsidered classic exemplars of H. erectus. But it is still worthwhile noting that as a group they dodiffer fairly markedly from the Trinil type material, though mostly in ways that recall theSambungmacan/Ngandong series. The crania from this site (with brain volumes ranging from850 to 1,140 ml) are most distinctive in their supraorbital morphology, with a low-set glabella,supraorbital tori having a strong vertical component, and a continuous posttoral sulcus. In addition,there are marked dental differences between the Zhoukoudian and Sangiran samples (Schwartz andTattersall 2005). Chinese specimens closely resembling the Peking Man materials include theNanjing crania (L€u and the Tangshan Archaeological Team 1996). Other materials sometimesassociated with them, such as those from Lantian, Hexian, Yunxian, and Longgupo, show a varietyof differences both from Zhoukoudian and among themselves. These differences are discussed bySchwartz and Tattersall (2005) and Tattersall and Schwartz (2009).

AfricaAs already noted, at one time or another many African fossils in the 1- to 2-myr time range havebeen referred to the species H. erectus. Among them, the classic exemplars are fossils from theTurkana Basin of northern Kenya, notable among these being the cranium KNM-ER 3733 and thecalvaria ER 3883 from sediments at Koobi Fora to the east of Lake Turkana, and the fairly completeskeleton KNM-WT 15000 from deposits to its west at Nariokotome. The mandible KNM-ER 992,initially allocated simply to Homo sp., comes from Ileret, to the north of Koobi Fora. All thespecimens concerned date within the 1.9- to 1.5-myr range. The widely used term “African Homoerectus” was a convenient designation for these fossils and a host of others, but it disguises the factthat a substantial variety of morphologies is involved.

This reality was first acknowledged in 1975 by Groves and Mazak who, as noted, made the ER992 mandible (Figs. 1 and 4), the holotype of the new species H. ergaster (“work man”).Subsequently, many authors have begun to use the new name in place of “African Homo erectus”to the extent that it is now H. ergaster that is the standard-issue Homo of the 2- to 1-myr period.Still, all this change has achieved is to remove the assortment of Asian morphologies from theAfrican equation, and it does nothing to address the morphological variety found within thecontinent in this period. In coming to grips with this, the best place to start is with the iconicexample, the KNM-ER 3733 cranium (Fig. 3), which has an endocranial capacity of 848 ml. In thisindividual, the supraorbitals arc separately over each orbit and project forward as well as upward.There is thus a distinct posttoral sulcus in front of the quite steep frontal rise, which rapidly peaksbefore the profile descends more gradually rearward. Seen from behind, the braincase is rather tallcompared to its breadth, and its side walls are curving. The raised temporal lines start quite farmedially. It is unsurprising that these characteristics distinguish this specimen sharply from anyAsian Homo; more remarkable are its differences from the ER 3883 cranium (Fig. 3), which has anendocranial volume of 804 ml. This individual has very thickened supraorbital margins thatprotrude outward but slightly down, overhanging nasion, and the (mostly missing) face beneath.Further, in this specimen the frontal slopes strongly up and back, reaching its maximum heightrather far back. Unlike in ER 3733, the mastoid is large and protruding, and in what is preserved of



of 18

the face the zygoma flares outward from top to bottom. That this morphology is no freak is shownby its close repetition in preserved features of the ER 3732 partial cranium (Fig. 3).

Interestingly, all the specimens just mentioned are distinctly different from the skull of theKNM-WT 15000 skeleton (Fig. 3), exhaustively described in the monograph edited by AlanWalker and Richard Leakey (1993). This 1.6-myr-old skeleton is remarkable both for its degreeof completeness and for being the earliest good evidence we have in the human fossil record of thearrival of essentially modern stature and postcranial proportions. Frustratingly, we cannot yet becertain that this was the case for the possessors of the ER 3733 and ER 3883 crania. The adolescentWT 15000 individual died at about 8 years of age, but was at a stage of development approximatingthat of a modern 12-year-old. This presumed male stood about 160 cm tall, but had he survived toadulthood, it is estimated that he would have topped 180 cm. He was long limbed and slender, withefficient heat-shedding proportions that would have served him well in the heat of the open tropicalsavanna. In contrast to the relatively long crania just described, the braincase of WT 15000, witha capacity of about 900 ml, was quite short and had a well-rounded profile. To the extent that it ispossible to judge, the badly damaged supraorbital surfaces amounted to little more than substantialthickenings of the superior orbital margins, lacking either the aggressive projection seen in ER3733 or the vertical thickening noted in ER 3883. The structure of the face contrasts with that seenin both ER 3733 and ER 3883; it is longer and narrower, with much more alveolar prognathism,a higher and narrower nasal aperture, and preserved portions of the nasal bones that suggest a flatterprofile of the upper face. It is often claimed that these differences from the East Turkana specimensare due to the subadult status of WT 15000, but this appears rather dubious since differences of thiskind would, if anything, probably have become more marked with age.

Cranial differences are backed up by dental comparisons to the extent that these are possible. ER3883 has no associated teeth, and ER 3733 has only one, an upper M2. But this tooth, thoughunfortunately quite heavily worn, is nonetheless distinctly different from its counterpart in WT15000. The upper M2 of ER 3733 has smooth enamel and well-defined trigon cusps, with theparacone much larger than the metacone. The hypocone is distally placed, the cristae are sharp, andboth basins are well excavated. InWT 15000, in contrast, the upperM2s are high 1/N crowned with

fairly flat but wrinkled occlusal surfaces. The cusps of the trigon are subequal in size, and the basinsare quite shallow. In both ER 3733 andWT 15000, the upperM2s contrast with their homologues inthe best-preserved upper dentition from Java, the Sangiran 4 palate, in which there are low cusps,a massive hypocone, and a very large postprotocrista that is not a feature of either Africanspecimen.

The ER 992 type specimen of H. ergaster is a lower jaw, so comparisons to ER 3733 and ER3883 are not possible. However, uniquely, the WT 15000 skull has a definitively associated lowerjaw (Fig. 4), allowing direct comparison to ER 992. When this comparison is made, cleardifferences become apparent. In ER 992 (Fig. 4), the lower canines are quite high and arecompressed buccolingually. In the anterior lower premolar, there are distinct anterior and posteriorfoveae, and the protoconid is the clearly dominant cusp. In the posterior premolar the protoconidand metaconid are subequal, and the basins are shallow. The elongated lower molars bear roundedand protruding hypoconulids, their basins are shallow, and their enamel is wrinkled. In contrast, thelower canines of WT 15000 are short crowned, with distinct mesial and distal foveae that bounda strong lingual pillar that swells out the tooth at its base. Both premolars have deep mesial anddistal basins. The first molar is distended mesially, and both erupted molars have large andlingually placed hypoconulids and narrow but deep talonid basins that are surrounded by well-defined but rather bulbous cusps. Here, again, we have two distinctly different lower dentalmorphologies, both of which also differ from their homologues known from Sangiran.



of 18

Morphologically, at least, the lower dentitions of ER 992 and WT 15000 are not “the same thing,”and both are at variance with Javan H. erectus (Schwartz and Tattersall 2000, 2005).

Interestingly, the 1.5-myr-old ER 992 from Ileret makes a fairly good match for the mandibleOH 22 (Fig. 4) from Olduvai Gorge in Tanzania, as well as for its rather older (1.9 myr) neighborER 3734 (Fig. 4) from Koobi Fora. As for WT 15000, its lower teeth compare quite closely withthose of OH 13, one of the paratypes ofH. habilis from Bed II of Olduvai Gorge, and of similar age.Both showmesially tapering premolars, with small metaconids lying opposite the protoconids, andsmaller foveae in front of these cusps than behind. The degree of wear on the molars is verydifferent, but both show an oblique groove that runs between the hypoconulid and hypoconid to thebase of the metaconid; and in both the hypoconid lies buccally and the M1s taper slightly distally,while the M2s are more broadly rounded at the rear. In the cranium and upper dentition, WT 15000shows substantial similarities with the East Turkana cranium ER 1813 (Fig. 3). Although the latterboasts a substantially smaller intracranial volume of 510 ml, both specimens share a short, highcranial vault (rather like that of OH 13) with rounded brows that arc over each orbit and a frontalthat rises behind a very short posttoral plane. In both the nasal apertures are tall, are relativelynarrow, and taper strongly upward, while the nasoalveolar clivuses are long and slope forward.

Even more telling are upper dental similarities amongWT 15000, ER 1813, and OH 13. All havea high-crowned but flat-surfaced M1, with the mesiodistally long hypocone as high as theprotocone, and separated from it by a lingual notch. In all, the M2s are basically similar to theMIs, but show reduction of the metacone and a smaller notch between the hypocone and protocone.In all, M1 and M2 both have thick postcingula. In WT 15000 and ER 1813, the P2s are very similarin having a bulbous and centrally placed paracone and a continuous crista running mesially fromthe paracone and swinging right around the side of the tooth. Further, although it has a very worndentition, the recently described palate OH 62 seems to present an upper dental morphology similarto those of the three specimens just described. In sum, the evidence seems to be quite compellingfor the existence of an upper dental morph, most spectacularly represented by the fairly completeindividualWT 15000, that is found almost a thousand kilometers away at Olduvai as well as aroundLake Turkana at Koobi Fora and Nariokotome. The available name for this morph, should anyone

wish to designate it a species, is H. microcranous (Ferguson 1995).Other “AfricanHomo erectus” specimens include the 1.4-myr-old OH 9 calvaria (Fig. 3: 967 ml)

from Olduvai Gorge, but it does not compare any better to the material from eastern Asia than itdoes to the Turkana fossils, and though it has been compared to the purportedly H. erectus Dakacranium fromEthiopia (�1.0myr, 995ml; Asfaw et al. 2002), resemblances between the two, otherthan in endocranial volume and its correlates, are not particularly striking (Schwartz and Tattersall2005). In sum, there is considerable morphological diversity among African Homo of the 2.0- to1.0-myr period; and this diversity does seem to be organized into a number of distinctive morphs.Some of these are represented by individual fossils such as OH 9 that are clear outliers in terms ofother known material; others seem to be represented by multiple individuals and even at multiplesites. Definitive systematic organization of this variety will clearly have to await a more compre-hensive fossil record, but it is already evident that we are not looking here at a chronologicaltransformation series, even one that is represented by high diversity at all time points. Somewherein all of this there is a systematic signal, and it is evident that the blanket appellation H. ergaster,while a useful device for distinguishing the African hominid radiation of this period from the Asianone, is an inadequate expedient for describing diversity in the African record.

Some paleoanthropologists are beginning implicitly to recognize this, but apparently withoutwishing to abandon old paradigms. Thus Spoor et al. (2007) reported new fossils from East Turkanathat they concluded were evidence for two contemporaneous lineages of Homo in the Turkana



of 18

Basin at around 1.5 million years ago. One of these was the calvaria KNM-ER 42700. Spoor andcolleagues cited a handful of trivial features that they claimed allied this specimen with Homoerectus; but in reality it bears none of the classic hallmarks of that species such as the long, low,posteriorly sharply angled lateral profile, the shelflike and protruding supraorbital surfaces, and theovoid posterior profile. Indeed, in these diagnostic features the Turkana specimen is the veryantithesis of the Javan type specimen. In contrast, the partial maxilla KNM-ER 42703, thoughhardly comparable to the calvaria, was considered to represent another lineage and was allocated tothe exceedingly poorly defined species Homo habilis. Whether or not that species assignment isappropriate, or even means anything at all, the allocation of ER 42700 to Homo erectus isextremely telling. For while Spoor and colleagues evidently accept diversity among 1.5-myr-oldhominids at Turkana, the assignment of the calvaria toHomo erectus only makes any sense at all inthe context of the notion that Homo erectus is the middle “grade” of a single comprehensive,worldwide, variable, and gradually evolving Homo lineage.

EuropeIt is now fairly widely accepted that most of the western European forms that have at one time oranother been described asH. erectus (see above) are better allocated toH. heidelbergensis, a movethat takes these fossils out of the scope of the current discussion. And while some still view theItalian Ceprano calvaria (ca. 800–900 kyr; 1,165 ml), recently designated the holotype of H.cepranensis by Mallegni et al. (2003), as a representative of H. erectus, nobody has claimed that

Fig. 5 The D 2600 mandible from Dmanisi. (a) Front view, (b) left lateral, and (c) occlusal (Photos courtesy of David

Lordkipanidze)



of 18

it represents H. ergaster. Indeed, Mounier et al. (2011) have suggested that it is a primitive Homoheidelbergensis.

To the east, however, in the Caucasus at the Georgian site of Dmanisi (1.8 myr), there existsa very important and early hominid fossil assemblage with claimed African affinities (Figs. 3, 4, 5,6, and 7). The first find, the mandible D 211 (Fig. 4) discovered in 1991, was assigned by itsdescribers to “archaic African Homo erectus” (Gabunia and Vekua 1995). It was more generallyattributed to H. erectus by Henke (1995) and also by Brauer and Schultz (1996), although theselatter authors remarked that, oddly, this early mandible showed “progressive” features seen ingeologically younger H. erectus. These early differences in interpretation foreshadowed a fairlywild taxonomic ride. Gabunia et al. (2000) reported the discovery of two crania (D 2280 and 2282;Fig. 3) close to the original site; they considered these comparable in size andmorphology to KoobiFora H. ergaster, although cranial volumes were somewhat smaller: 780 and 650 ml, respectively.With the discovery in September 2000 of a very large and long mandible with highly worn teeth (D2600; Fig. 5), the picture changed again. This mandible presented a marked contrast to D 211, but

Fig. 6 Five views of the D 2700 cranium from Dmanisi (From Vekua et al. (2002), supplemental web data; Photos

courtesy of David Lordkipanidze)



of 18

the Dmanisi team nonetheless concluded that all of the specimens belonged to a single highlysexually dimorphic species which they namedH. georgicus, with D 2600 as its holotype (Gabouniaet al. 2002). The gracile specimens D 211, D 2280, and D 2282 were considered to be female; andthe robust D 2600 lower jaw was viewed as a male representative of this species. The groupconcluded that H. georgicus “preserves several affinities with Homo habilis and Homo rudolfensis. . . foretelling the emergence of Homo ergaster” (Gabounia et al. 2002, p. 245). But things did notstop there. Almost simultaneously, the team announced the discovery of an associated cranium andmandible, D 2700/D 2735 (Vekua et al. 2002; Fig. 6). Strikingly different from the craniadiscovered earlier, though also notably small-brained (600 ml), this excellently preserved speci-men was said by its describers to bear resemblances to penecontemporaneous East African fossils.

Fig. 7 The edentulous D 3444 cranium from Dmanisi. Lower left: mandible D 3900. Lower right: CT-based

superimpositions by C. Zollikofer and M. Ponce de Leon of the Dmanisi skull D 3444/D 3900 on the D 2700/D

2735 and D 2282/D 211 specimens, to show contrasting silhouettes (From Lordkipanidze et al. (2005). Photos courtesy

of David Lordkipanidze. Scale: 10 cm)



of 18

Abandoning the species H. georgicus, as well as the notion of H. ergaster as a separate entity, theDmanisi team allocated D 2700/D 2735, and the rest of the hominid assemblage along with it, toH.erectus, while noting that they “are among the most primitive individuals so far attributed” to thatexpanded species (Vekua et al. 2002, p. 88).

During the 2002/2004 field seasons yet another associated cranium and mandible (D 3444/D3900) were recovered at Dmanisi. The most remarkable aspect of this aged presumed male (Fig. 7)is that he had possessed just a single tooth at death and had evidently been largely edentulous formany years (Lordkipanidze et al. 2005). Although at least one recent chimpanzee is known to havesurvived a long time in an edentulous state, the Dmanisi team surmised that the individual musthave “survived for a lengthy period without consuming foods that required heavy chewing . . . and/or by virtue of help from other individuals” (Lordkipanidze et al. 2005, p. 718), and suggested thatthis had significant implications for early hominid social structure. They also noted that thecranium had been found in close proximity to Mode 1 stone artifacts and to cut-marked animalbones. The authors refrained from commenting on the systematic implications of the new find; butin a review published soon afterward, Rightmire et al. (2006) reaffirmed their belief that theDmanisi assemblage as a whole was a single “paleodeme” best placed within H. erectus (whichto them subsumed H. ergaster), despite resemblances to H. habilis in brain volume and in someaspects of craniofacial morphology. Yet while they (Rightmire et al. 2006, p. 140) noted that, if thelarge D 2600 mandible could be accommodated within the rest of the Dmanisi hominid population,then “the appropriate nomen isH. erectus georgicus,” they also pointed out that should the separatespecies status advocated by Gabounia et al. (2002) for the large jaw be “verified by new discov-eries, then a subspecies other than H. erectus georgicus will have to be selected [for the remainderof the sample].”

The hoped-for new discovery, of the cranium matching the D 2600 jaw, was duly made atDmanisi in 2005. Following a long agony of indecision, this complete and beautifully preservedspecimen (D 4500) was finally published in 2013 (Lordkipanidze et al. 2013). The delay is quiteunderstandable, for this fossil is quite unlike anything ever seen before. As one would havepredicted from the large teeth and long tooth rows of the mandible, the new cranium possesses

a large and highly prognathic face, which is hafted on to a remarkably small braincase witha volume of fractionally less than 550 ml. But while these attributes generally recall those ofaustralopiths, Lordkipanidze and colleagues probably wisely refrained from associating their newspecimen with any previously known early hominid taxon. Instead, they crammed it into Homoerectus, an assignment for which there is less than scant morphological justification. Declaring thatthe extraordinarily distinctive morphology of D 4500 “reflects variation between demes of a singleevolving lineage” (Lordkipanidze et al. 2013, p. 330), the group concluded that “specimenspreviously attributed to H. ergaster are thus sensibly classified as a chronosubspecies, H. erectusergaster” (Lordkipanidze et al. 2013, p. 330). And finally they took the astonishing step of reducingthe entire Dmanisi hominid sample to the previously unknown status of a sub-subspecies, Homoergaster erectus georgicus. If there is any substance whatever to this convoluted taxonomicjudgment, systematists everywhere should be in mourning, because it effectively deprives mor-phology of any utility in systematics – and where are we left to go from there?

In any event, the contorted taxonomic journey of the Dmanisi hominids reflects the unusualmorphologies that make it hard to fit them into established categories. And this journey is certainlynot yet at an end. Even before the initial discovery of D 4500, Schwartz and Tattersall (2005) hadalready noted that the morphological heterogeneity in the Dmanisi assemblage made it difficult torecognize a single consistent morph at the site, irrespective of what this might have implied aboutspecies status. What is more, the lead geologist at Dmanisi, Reid Ferring (personal



of 18

communication), reports that accumulation of the Dmanisi hominids might have taken place overthe span of as much as a few hundred years. And if this is the case, there is no compelling reason (asthere might be in the case of a catastrophic assemblage) to believe that only one hominid species isnecessarily implicated in the fossil assemblage. Additionally, Schwartz and Tattersall observedthat none of the Dmanisi material appeared to bear very close, i.e., systematically suggestive,resemblances either to any Asian fossils that had been described as H. erectus or to any Africanspecimens allocated toH. erectus or toH. ergaster. Exactly howmuch systematic variety there is inthis assemblage clearly awaits more study; but although the Dmanisi hominids most plausiblyrepresent one or possibly more early departures from Africa hard on the heels of the origin ofHomo, it is hard at present to point to craniodental morphologies that specifically unite them withany latest Pliocene or earliest Pleistocene African hominids yet known.

It is unknown with any certainty exactly what it was that allowed hominids to spread, for(presumably) the first time, out of the ancestral continent of Africa as far as Dmanisi. But sincetheMode 1 stone tool assemblage at Dmanisi is remarkably primitive, and the brains of the hominid(s) there remained small, it seems most likely that the factor responsible was the acquisition ofstriding bipedality. As exemplified by WT 15000, this locomotor style was already present inHomo ergaster; and a partial postcranial skeleton from Dmanisi (putatively associated with theD 2700/D 2735) and postcranial bones from three other individuals are said to show “modern-human-like body proportions” (Lordkipanidze et al. 2007, p. 305). Still, the scientists whodescribed these elements also noted a “surprising mosaic of primitive and derived features”(Lordkipanidze et al. 2007, p. 305); and it is clear that the last word on the Dmanisipostcranials – and indeed, on the assemblage as a whole – has yet to be written.

Conclusions

Homo ergaster is the designation of choice for the growing number of paleoanthropologists whobelieve that the fossils previously allocated to “African Homo erectus” are sufficiently different

from the Asian type material of H. erectus to warrant assignment to a distinct species. Adoption ofthis nomenclature is a considerable improvement on our older understanding, certainly to theextent that it emphasizes that the Trinil fossil and others like it are quite highly autapomorphic andthat Javan or at least eastern Asian H. erectus is thus most appropriately viewed as an indigenousand terminal regional species (or maybe even clade), rather than as an Old World-wide stage orgrade in hominid evolution. Nonetheless, it remains true that to call the entire African rump of thisOld World “Homo erectus grade” H. ergaster, is to brush a huge diversity of morphologies underthe rug of one single species. Close examination of the morphologies displayed by the diversity offossils that have at one time or another been referred to as “AfricanHomo erectus” makes it evidentthat, while it is likely that all may be legitimately regarded as members of a single hominid clade,there is some diversity within it.

The KNM-ER 992 holotype mandible of H. ergaster appears to be matched morphologically bythe OH 22 mandible from Olduvai as well as by another mandible (ER 3734) from Koobi Fora. Butthere seems to be no compelling reason to match these lower jaws with any of the cranial materialsavailable, and certainly none to associate themwith the iconic KNM-WT 15000 skeleton, the lowerdentition of which is distinctly different in a whole host of characteristics. Clearly, there is a needfor a detailed systematic reappraisal of the entire “AfricanHomo erectus”¼Homo ergaster group.Meanwhile, the recognition of a distinct H. ergaster clade at least serves to highlight the fact thatthe complexity of the hominid evolutionary story throughout the Old World in the Early–Middle



of 18

Pleistocene was far greater than is implied by the inclusion of the entire group of fossils involvedwithin the single hugely variable species H. erectus.

References

Asfaw B, Gilbert WH, Bayene Y, Hart WK, Renne PR, WoldeGabriel G, Vrba ES, White TD(2002) Nature; 416:317–20.

Boule M. L’homme fossile de la Chapelle-aux-Saints. Ann Paleontol. 1911–1913;6:1–64,7:65–208, 8:209–79.

Brauer G, Schultz M (1996) The morphological affinities of the Plio-Pleistocene mandible fromDmanisi, Georgia. J Hum Evol 30:445–481

Cunningham D (1895) The place of “Pithecanthropus” on the genealogical tree. Nature 53:269Ferguson WW (1995) A new species of the genus Homo (Primates: Hominidae) from the Plio-

Pleistocene of Koobi Fora, in Kenya. Primates 36:69–89Gabounia L, de Lumley M-A, Vekua A, Lordkipanidze D, de Lumley H (2002) Decouverte d’un

nouvel hominide a Dmanissi (Transcaucasie, Georgie). C R Paleovol 1:243–253Gabunia L, Vekua A (1995) A Plio-Pleistocene hominid from Dmanisi, East Georgia, Caucasus.

Nature 373:509–512Gabunia L, Vekua A, Lordkipanidze D, Swisher CC, Ferring R, Justus A, Nioradze M,

Tvalcrelidze M, Anton S, Bosinski GC, Joris O, de LumleyMA,Majsuradze G, MouskhelishviliA (2000) Earliest Pleistocene hominid cranial remains from Dmanisi, Republic of Georgia:taxonomy, geological setting and age. Science 288:1019–1025

Groves CP, Mazak V (1975) An approach to the taxonomy of the Hominidae: GracileVillafranchian hominids of Africa. Cas Miner Geol 20:225–247

Henke W (1995) Qualitative and quantitative analysis of the Dmanisi mandible. In: Radlanski RJ,Renz H (eds) Proceedings of the 10th international symposium on dental morphology. Christineund Michael Br€unae GbA, Berlin, pp 459–464

Holloway RL, Broadfield DC, Yuan MS (2004) The human fossil record. Vol. 3: brain endocasts,the paleoneurological evidence. Wiley-Liss, New York

Leakey R (1972) Further evidence of lower Pleistocene hominids from East Rudolf, North Kenya.Nature 237:264–269

Leakey R, Walker A (1976) Australopithecus, Homo erectus and the single species hypothesis.Nature 261:572–574

Lordkipanidze D, Vekua A, Ferring R, Rightmire GP, Agusti J, Kiladze G, Mouskhelishvili A,Ponce de Leon M, Tappen M, Zollikofer CPE (2005) The earliest toothless hominin skull.Nature 434:717–718

Lordkipanidze D, Jashashvii T, Vekua A, Ponce de Leon M, Zollikofer CPE, Rightmire GP,Pontzer H, Ferring R, Oms O, Tappen M, Bukhsianidze M, Agusti J, Kahlke R, Kiladze G,Martinez-Navarro B, Mouskhelishvili A, Nioradze M, Rook R (2007) Postcranial evidence ofearly Homo from Dmanisi, Georgia. Nature 449:305–310

Lordkipanidze D, Ponce de LeonM,Margvelashvili A, Rak Y, Rightmire GP, Vekua A, ZollikoferCPE (2013) A complete skull from Dmanisi, Georgia, and evolutionary biology of early Homo.Science 342:326–331

L€u Z, The Tangshan Archaeological Team (1996) Locality of the Nanjing man fossils. CulturalRelics Publishing House, Beijing



of 18

Mallegni F, Carnieri I, Bisconti M, Tartarelli G, Ricci S, Bidittu I, Segre AG (2003) Homocepranensis sp. nov. and the evolution of African-European Middle Pleistocene hominids.C R Paleovol 2:153–159

Mayr E (1950) Taxonomic categories in fossil hominids. Cold Spring Harb Symp Quant Biol15:109–118

Mounier A, Condemis S, Manzi G (2011) The stem species of our species: A place for the aschaichuman cranium from ceprano, Italy. PLoSOne 6(4):e18821

Rightmire GP (1990) The evolution ofHomo erectus: comparative anatomical studies of an extincthuman species. Cambridge University Press, Cambridge

Rightmire GP, Lordkipanidze D, Vekua A (2006) Anatomical descriptions, comparative studiesand evolutionary significance of the hominin skulls from Dmanisi, Republic of Georgia. J HumEvol 50:115–141

Schwalbe G (1899) Studien €uber Pithecanthropus erectus Dubois. Morphol Anthropol 1:16–228Schwartz JH, Tattersall I (2000) What constitutes Homo erectus? Acta Anthropol Sin

19(Suppl):18–22Schwartz JH, Tattersall I (2003) The human fossil record. Vol. 2: craniodental morphology of

genus Homo, Africa and Asia. Wiley-Liss, New YorkSchwartz JH, Tattersall I (2005) The human fossil record. Vol. 4: craniodental morphology of early

hominids (genera Australopithecus, Paranthropus, Orrorin) and overview. Wiley-Liss, NewYork

Spoor F, Leakey MG, Gathogo PN, Brown FH, McDougall I, Kiarie C, Manthi KF, Leakey LN(2007) Implications of new early Homo fossils from Ileret, east of Lake Turkana, Kenya. Nature448:688–691

Tattersall I (1986) Species recognition in human paleontology. J Hum Evol 15:165–175Tattersall I (1992) Species concepts and species identification in human evolution. J Hum

Evol 22:341–349Tattersall I (1995) The fossil trail: how we know what we think we know about human evolution.

Oxford University Press, New YorkTattersall I (1998) Becoming human: evolution and human uniqueness. Harcourt Brace, New YorkTattersall I (2008) The fossil trail: how we know what we think we know about human evolution,

2nd edn. Oxford University Press, New YorkTattersall I, Schwartz JH (2009) Evolution of the genusHomo. Ann Revs Earth Planet Sci 37:67–92Theunissen B (1988) Eugene Dubois and the ape-man from Java: the history of the missing link and

its discoverer. Kluwer, DordrechtVekua A, Lordkipanidze D, Rightmire GP, Agusti J, Ferring R, Majsuradze G, Mouskhelishvili A,

Noradze M, Ponce de Leon M, Tappen M, Tvalchrelidze M, Zollikofer C (2002) A new skull ofearly Homo from Dmanisi, Georgia. Science 297:85–89

Walker A, Leakey R (1978) The hominids of East Turkana. Sci Am 239(2):44–56Walker A, Leakey R (1993) The Nariokotome Homo erectus skeleton. Harvard University Press,

Cambridge



of 18

Index Terms:

Homo ergaster_3–5, 8, 11Homo ergaster

African_8Homo ergaster

Asian_5Homo ergaster

cranial morphologies_4Homo ergaster

European_11Homo ergaster

existence_3



of 18

Homo ergaster and Its Contemporaries … · systematics and taxonomy of the genus Homo in the Early...

Documents

Transcript of Homo ergaster and Its Contemporaries … · systematics and taxonomy of the genus Homo in the Early...