Modern human origins

YEARBOOK OF PHYSICAL ANTHROPOLOGY 32135-68 (1989)

Modern Human Origins

FRED H. SMITH, ANTHONY B. FALSETTI, AND STEVEN M. DONNELLY Department of Anthropology, University of Tennessee, Knoxuille, Tennessee 37996-0720

KEY WORDS piens

Paleoanthropology, Archaic Homo sapiens, Modern Homo sa-

ABSTRACT During the past decade or so, considerable new data pertinent to the origin of modern humans have come to light. Based on these new data and reinterpretation of older information, three models have been offered to explain the development of modern people. These models-Brau- er’s Afro-European sapiens hypothesis, Stringer and Andrew’s recent Afri- can evolution model, and Wolpoff, Wu, and Thorne’s multiregional evolution model-have their roots in earlier models but differ from most by virtue of their worldwide perspective and integration of genetic and paleoanthropological data pertinent to modern human origins. This review presents a detailed discussion of these data in light of the three models. While convincing arguments can be offered for each of these models, it is concluded that none are unequivocally supported by the available data.

I confess that I cannot recall any case within my experience which looked at first glance so simple, and yet which presented such difficulties (Sherlock Holmes-The Man with the Twisted Lip).

The issue of modern human origins is the oldest controversy in the controversy- laden discipline of human paleontology. In fact, the questions of when, where, and especially how people fundamentally similar to ourselves first appeared were sub- jects of considerable debate for natural historians, philosophers, and theologians long before the existence of a human fossil record was documented. Despite the antiquity of the issue, both interest in and debate on it are no less intense today than they were following the recognition of the first human fossil relevant to the controversy in Germany’s Neander Valley 133 years ago. The strength of current interest is perhaps most clearly reflected by the fact that, since 1984, no less than four major scientific volumes have been, or will be, devoted to discussion of modern human origins (Smith and Spencer, 1984; Mellars and Stringer, 1989; Trinkaus, 1989; Brauer and Smith, 1990). Although they evince a notable lack of unanimity in terms of interpretation, the papers published in these volumes and elsewhere during the last decade reveal that existing knowledge pertinent to the evolution of modern people is impressive. In addition to an ever-expanding fossil record and a large number of new analytical techniques that facilitate study of this record, various forms of genetic data on living human populations and a series of new chronometric age estimates for key hominid fossils are profoundly influencing scientific perspectives on how we came to be as we are. The purpose of this review is to discuss these new data and the major explanatory models derived from them, which have accumulated largely during the past decade. The review will focus on

0 1989 Alan R. Liss. Inc

36 YEARBOOK OF PHYSICAL ANTHROPOLOGY [Vol. 32, 1989

new developments in the human fossil record, its chronological framework, and molecular biology. It will not systematically consider earlier perspectives on modern human origins, since these are dealt with elsewhere (Spencer, 1984, 1986; Spencer and Smith, 1981; Trinkaus, 1982). Furthermore, detailed assessment of the archaeological record and the various behavioral inferences drawn therefrom are beyond the scope of this review.

MODELS OF MODERN HUMAN ORIGIN

As recently as the early 1980s (e.g., Spencer and Smith, 19811, the issue of modern human origins was still being examined in the context of the three models identified by Vallois (1958): the Neandertal phase, Presapiens, and Preneandertal models. The Neandertal phase model is traceable to the unilineal schemes proposed by Schwalbe, GorjanoviC-Kramberger, and HrdliEka (RadovEiC, 1988; Spen- cer, 1984) and claimed that European Neandertals represented an ancestral stage in the phylogeny of modern humans. The Presapiens hypothesis, as developed by Boule (1921), posited the existence of a European lineage leading to modern people but separate from that of Neandertals. The Preneandertal model regarded gener- alized Neandertals as logical ancestors for both the “Classic” Western European Neandertals and modern humans but asserted that the latter evolved outside Europe and were not lineal descendants of the former. This perspective is best presented in classic papers by Howell (1951, 1957) and, more recently, Howells (1976).

These models were formulated at a time when studies of later hominid evolution were characterized by a distinct Eurocentrism. This was an understandable situation for several reasons, principally the relative lack of knowledge available from other regions (Smith, 1985; Spencer, 1984; Trinkaus, 1982). However, as the pertinent homind fossil record expanded and as archaeological and chronological frameworks became more reliable for regions other than Europe, the situation began to change. The seeds of this change could be seen in the 1970s, particularly in the insightful essay by Howells in 1976, but it has been mainly in this decade that primary attention has been focused on regions other than Europe in the quest to explain the emergence of modern people.

Since 1980, three “new” models have been proposed to account for the development of modern humans. All of these models have their roots in earlier models and thus are not entirely new. However, they differ from earlier models in that each emphasizes fossil samples from areas other than Europe, takes a worldwide perspective on the issue of modern human origins, and incorporates present-day knowledge of population dynamics and genetics of extant human populations into its assessment of this issue.

The A fro-European sapiens hypothesis The Afro-European sapiens (AES) model was formulated by Gunter Brauer

(1984a-c) on the basis of his extensive research on African Late Pleistocene/Ho- locene skeletal series (1978, 1983b) and the pertinent Middle and Late Pleistocene African fossil record. Although certain elements of the AES model can be seen in the work of others (notably Rightmire 1979, 1984, 19861, Brauer was the first to argue systematically for an early, indigenous transition to modern Homo sapiens in Africa and a subsequent spread of this new human form into Europe (1984a) and, perhaps with less certainty, Asia as well (1984b).

Brauer asserts that modern humans evolved from an early archaic H. sapiens stage, primarily represented by hominids from the sites of Bod0 (Ethiopia), Kabwe (Zambia), Ndutu and Eyasi (Tanzania), and Elandsfontein (Republic of South Af- rica-RSA). From this primitive, rather erectus-like ancestral H. sapiens group emerged a late archaic H. sapiens stage, which was transitional between the Bodol Kabwe group and early modern Africans. The key representatives of the latter group are the Ngaloba (Tanzania), Florisbad (RSA), and Omo 2 (Ethiopia) crania. Recently Brauer has suggested that, on morphological grounds, both the Eliye

Smith et al.1 A REVIEW OF MODERN HUMAN ORIGINS 37

Springs and Guombe crania (Kenya) also belong with this series (Brauer and Leakey, 1986; Brauer, 1989). He argues that available chronological information establishes an age of >lo0 ky ago for both the early and late archaic H . sapiens stages and presents an extensive series of morphometric analyses to demonstrate the transitional nature of the late archaic H. sapiens (transitional) group (Brauer, 1983a, 1984a). Modern humans, according to Brauer, are present in southern Africa approximately 100 ky ago at sites like Border Cave and Klasies River Main Site (RSA), Omo Kibish site KHS (Ethiopia), and possibly Mumba Rock Shelter (Tanzania). Furthermore, he accepts recent mitochondrial (mt) DNA (Cann et al., 1987; Stoneking and Cann, 1989) and nuclear DNA (Nei, 1985; Wainscoat et al., 1986) analyses as strong evidence of an African ancestry of modern humans (Brauer, 1989).

Brauer sees the movement of modern populations out of southern Africa as a gradual process, probably brought on at least in part by climatic and environmental desiccation in Africa (Brauer, 1984a; 1989). However, he does not view modern humans as the result of a biological speciation event. Rather he suggests that the expanding modern populations assimilated some archaic genes into their gene pools, and therefore that the disappearance of archaic human populations in Eur- asia was the result of “replacement and hybridization.” This process, according to Brauer (1989) “. . . was certainly complex, multicausal, different in various regions and hardly rapid or complete.”

Thus, the AES model would recognize that some degree of local continuity between archaic and modern humans was possible anywhere in Eurasia, and Brauer has used this argument to explain the Neandertal-reminiscent features of the Hahnofersand frontal (Brauer, 1980), as well as those noted by Smith (1982,1984), for early modern central Europeans. However, by far the dominant factor responsible for the emergence of modern Eurasians was, in Brauer’s view, the migration of African-derived populations into these parts of the Old World, not local continuity. Consequently, Brauer regards examples of diachronic change toward the modern human condition in Eurasian archaic groups as either parallelisms or the result of studies based on biased samples.

Recent African evolution model In many ways the recent African evolution (RAE) model is similar to the AES

model. It posits a recent (within the last 200 ky) African origin for all modern humans, suggests that the transition to modern humans occurred only in Africa, and claims that modern humans appeared in Africa earlier than anywhere else. It is clear that in developing this model, Stringer and Andrews (1988) drew exten- sively on the paleontological interpretations of Brauer, as well as Rightmire, in their assessment of the African fossil record. However, the RAE model appears to differ in one very significant point from the AES model. Stringer and Andrews’ presentation of the model in Science strongly implies that the African origin of modern humans was a biological speciation event. Therefore, once modern humans emerged from their archaic ancestors in southern Africa and began to radiate into Eurasia, no significant admixture would be possible between them and the archaic Eurasians they encountered, because they would have been different biological species. Two assertions made by Stringer and Andrews are particularly significant in the context of this implication. First, according to Stringer and Andrews, there is little or no evidence for local continuity across the archaiclmodern human boundary outside Africa. Second, any archaic-reminiscent traits in Eurasian populations are either primitive retentions found in the incoming African-derived populations or homoplasies.

It is also noteworthy that, while the molecular data have been cited as support for the AES model, they were not crucial in its formulation. For Stringer and Andrews, however, the indication of low nuclear and mitochondrial genetic variability in humans was a critical, if not the key, factor in the formulation of the RAE model. Although Stringer has clearly long favored a classic monocentric


origin for modern humans and the replacement of Eurasian Neandertals (Stringer, 1978, 1982; Stringer et al., 19841, his focus on Africa as the modern human homeland and the emphasis on speciation seemingly coalesced as a result of the mitochondrial DNA analyses of Cann et al. (1987). Cann and her colleagues argued strongly that all modern human mtDNA had a common ancestry in Africa between 140 ky and 290 ky ago, that the ancestral stock of modern Eurasians diverged from the African stock between 90 ky and 180 ky ago, and that interbreeding between these modern people and indigenous Eurasian archaic hominids was, at most, minimal. In fact, in a more recent assessment of the implications of their mtDNA data on Late Pleistocene human population history, Stoneking and Cann (1989) state that “. . . the rather staggering implication is that the dispersing African population [of modern humans] replaced the non-African resident populations [Eurasian archaic H . sapiens] without any interbreeding.”

Based on their interpretation and presentation of the paleoanthropological and genetic evidence, Stringer and Andrews’ RAE model has been characterized as a total replacement or speciation/replacement model (Smith, 1990; Smith and Pa- quette, 1989; Wolpoff, 1989; Wolpoff et al., 1988). Recently, Stringer (1990) has stated that his definition of H. sapiens is rooted in an evolutionary, not biological, species concept and is based on morphological grounds (synapomorphies in recent skeletal and early modern fossil remains). He also argues that some hybridization could occur between Neandertals and anatomically modern humans without their being conspecific. This would appear to remove the main point of difference between the RAE and AES models. However, Stringer also states that evidence of hybridization between Neandertals and modern humans has not been established. From our perspective, the effect of the RAE model has been to promote thinking of the origin of modern humans as a biological speciation event and thus that the fate of Neandertals and other archaic Eurasians was extinction without issue. This effect is quite well reflected by this recently published popular assessment:

If Neanderthal behavior was as relatively rudimentary and Neanderthal anat- omy as distinctive as I suspect, few Cro-Magnons may have wanted to mate with Neanderthals. And if Neanderthal women were geared for a 12-month preg- nancy, a hybrid fetus might not have survived. My inclination is to take the negative evidence at face value, to accept that hybridization occurred rarely if ever, and to doubt that any living people carry any Neanderthal genes. (Dia- mond, 1989:59)

The speciation/replacement aspect of the RAE model has perhaps been emphasized more by commentators on the Stringer and Andrews Science article than by the authors themselves (Delson, 1988; Gould, 1988, 1989; Lewin, 1988), because such an interpretation fits neatly certain perspectives on how evolution operates. Since the 1970s, some paleontologists have asserted that all macroevolution (origin of new species and higher taxa) is the result of cladogenesis (the punctuated equilibrium model) and that hominid evolution is no exception to this pattern (Eldredge and Tattersall, 1982; Gould, 1987, 1989; Stanley, 1979). Obviously, if modern humans resulted from cladogenesis in southern Africa, then radiated as a new species and replaced Eurasian archaic hominids, modern human origins could be cited as yet another example confirming the ubiquity of this process in macroevolution. Thus, much of the critical acclaim the REM model has received, particularly outside anthropology, stems from its speciation/replacement leanings. Re- gardless of whether Stringer and Andrews intended RAE to be interpreted as a total replacement model, it certainly has been widely interpreted as such and will be discussed in this context in this review.

The multiregional evolution model The multiregional evolution (MRE) or regional continuity model was first out-

lined in a broad theoretical context by Wolpoff et al. in 1984, although critical

Smith et al.] A REVIEW OF MODERN HUMAN ORIGINS 39

aspects of the model are discussed in earlier publications (Thorne, 1981; Thorne and Wolpoff, 1981; Wolpoff, 1980). Basically, the multiregional model argues that there is considerable morphological and genetic continuity across the archaidmodern human boundary in regions throughout Eurasia, as well as Africa. This would mean that paleontological indications of continuity, particularly transitional fossils and regionally distinct morphological features linking archaic and modern humans in specific geographic regions, should be found in Eurasia as well as Africa, but not necessarily that such indications should be common (Wolpoff et al., 1988) or that significant continuity characterizes every region. The timing and mode of appearance of modern humans (or modern human anatomical form) in a particular region, and the degree of local continuity involved, would result from a complex interplay of factors (the primary ones being the pattern of local selection and the extent and pattern of gene flow into that region), not simply from population replacement. Since both the AES and RAE models emphasize replacement, either with or without “hybridization,” as the key factor in the appearance of modern peoples outside Africa, the MRE model differs fundamentally from both. Clearly, then, if any type of speciation is involved in modern human origins, the multiregional model would predict something like what Van Valen (1986) has termed geographically extended anagenesis, not cladogenesis and replacement, as the responsible mode.

The roots of the MRE model are to be found in Weidenreich‘s polycentric views of modern human emergence, and as was also true for Weidenreich, the Middle and Late Pleistocene fossil human record from East Asia has played a pivotal role in the modern version of MRE (Thorne and Wolpoff, 1981; Wolpoff 1985, 1989, 1990; Wolpoff et al., 1984). Also like Weidenreich, modern supporters of multiregional evolution trace differences in a number of specific regional morphological complexes to ecological and demographic factors associated with the initial radiation of Homo erectus, the so-called center-and-edge perspective (Wolpoff et al., 1984). Un- like other polycentric models that have also been derived from Weidenreich’s, for example that of Coon (1962), the present-day MRE model does not claim that local lineages leading to modern humans evolved independently from each other. Rather, proponents of the MRE model assert that gene flow, in the form of both population movement (migration) and genetic exchange across population boundaries, would have prevented speciation between the regional lineages and thus maintained human beings as a single, although obviously polytypic, species throughout the Pleistocene.

The most ardent defender of MRE, Milford Wolpoff (1990) has listed a number of other predictions that he feels emerge from the multiregional model. First, he argues that no single definition of modern humans will apply in all regions because of the different mixtures of indigenous and extraneous factors that occurred in different geographical regions (see also Wolpoff, 1986). Second, modern human anatomical form does not necessarily have to appear earliest in Africa. Third, the earliest modern people in Eurasia will lack distinctly African regional features, since they are not the result of migrating Africans. Fourth, non-adaptive or regional clade features will be identifiable earliest in the most peripheral regions of the human range (before the appearance of modern humans) and late in the central region. Finally, Wolpoff and others (Excoffier and Langaney, 1989; Smith, 1990; Smith and Paquette, 1989; Spuhler, 1988, in press; Wolpoff, 1989, 1990; Wolpoff et al., 1988) have pointed out that there are explanations for extant human genetic variability that do not invoke a recent African origin for all modern people.

The species concept i n modern human origins One of the major issues in modern biology is the question of which species

concept to employ (e.g., Levinton, 1988; Mayr, 1969, 1982; Stanley, 1979; Wiley, 1981). Mayr’s biological species concept defines species on the basis of reproductive isolation and the possession of adaptations that allow coexistence with potential competitors (Mayr, 1963). This concept has enjoyed wide acceptance, but certainly


becomes problematic when applied to fossil samples. Mayr (1982:273) himself has written that “the biological species concept . . . is meaningful and truly applicable only in the nondimensional situation. It can be extended to multidimensional situations only by inference.” However, Mayr argues that the protected gene pool derived from reproductive isolation results in a discontinuity of ll. . . morphology and other aspects of the phenotype. . .” between species (1969:28) and thus that the amount and kind of morphological difference between samples may be used as evidence to draw inferences about reproductive isolation. The crucial problem in paleontological samples, of course, is to determine just what constitutes sufficient kind and amount of morphological differences to infer the existence of reproductive barriers and hence separate species.

Mayr (1963) as well as other proponents of the biological species concept, certainly recognize that hybridization between recognized, established living species occurs; but many cases of secondary contact and hybridization would presumably constitute evidence of incomplete speciation (see also Wiley, 1981:27) in both ne- ontological and paleontological contexts. In such instances, it is not certain that referring to the “populations” involved as separate species would be advisable. Furthermore, it can certainly be asserted that speciation can occur phyletically, as well as through cladogenesis (Mayr, 1963,1969,1982; Levinton, 1988; Van Valen, 1986). According to Levinton (1988:408), “. . . a large number of studies demonstrate the common occurrence of phyletic evolution of a sufficient magnitude to produce change on the order of specific and generic differences.” The occurrence of incomplete speciation can, in our opinion, be viewed as an aspect of phyletic change, potentially contributing to geographically extended anagenesis.

An alternative species concept, widely held among cladists, is the evolutionary species. Wiley (1982:25) defines an evolutionary species as “. . . a single lineage of ancestor-descendant populations that maintains its identity from other such lineages and that has its own evolutionary tendencies and historical fate.” With this definition, a lineage between branching points cannot be subdivided into different species, thus speciation can only occur by cladogenesis. The concept of paleospecies or morphospecies (cf. Simpson, 1961) is considered too arbitrary to reflect biological reality. A corollary to this perspective is that significant evolutionary change takes place only during the process of speciation. Thus, a connection is asserted between speciation and morphological change; the latter does not occur without the former.

The evolutionary species concept certainly does not deny the importance of re- production to the delineation of species. Wiley (1981), for example, notes that species must be reproductively isolated to the extent that they maintain their separate identities, evolutionary tendencies, and historical fates. He recognizes that the degree of introgression in zones of sympatry between different geographic populations is critical to determining if they constitute separate species.

Using an evolutionary species concept, Tattersall (1986; see also Lewin, 1989) has argued that Neandertals clearly represent a different species than modern humans and that still other species should be formally recognized in the Middle and Late Pleistocene hominid fossil record. Stringer (1990) has also stated that his separation of Neandertals and modern humans at the species level is based on the evolutionary species concept. In addition, he states that hybridization cannot be used to examine the relationships between Neandertals and modern humans.

Neandertals certainly maintain a degree of distinct identity compared to southern African and East Asian contemporaries, and there are unquestionably distinct tendencies that characterize Neandertal evolution. Thus, a case can certainly be made for separating H . sapiem and H. neanderthalensis within the framework of the evolutionary species concept. But, i t can be argued that evidence of gene flow between Neandertals and other populations (in the form of Neandertal features in early modern samples) means that Neandertals contributed, potentially exten- sively, to modern human gene pools in Europe and western Asia. If this can be established, assigning Neandertals to a separate species and attaching assump-


tions that this lineage became extinct without issue would make little biological sense. Wiley (1981:63) has stated that even if each population is a distinct epiphe- notype in allopatry, the populations are probably part of a single species if “introgression occurs over a wide geographic area and the populations are nearest rel- atives.” We opt, in this review, to use terms such as archaic humans and modern humans or archaic and modern H. sapiens, because we are not convinced that some of the implications associated with using H. sapiens and H. neanderthalensis are unequivocally valid. We remain unsure that the evolutionary species concept is really an improvement over the biological species concept, particularly that the former concept has greater biological validity than the latter. Levinton (1988), for example, has argued that cladogenesis has not been commonly demonstrated in the fossil record and that there is only very limited evidence for the association of morphological evolution with speciation. However, we do not intend to imply that Neandertals and modern humans are demonstrably not separate species. Tatter- sall (quoted in Lewin, 1989:1666) has chided that many anthropologists “. . . fail to come to grips with the biological realities surrounding the origins of modern humans.” We assert that, despite the convictions of some, exactly what the “biological realities” are is not yet clear.

AFRICA-THE SEARCH FOR EDEN

Nowhere has the interpretation of Late Pleistocene human evolution changed more radically in recent years than in Africa. As recently as the early 1970s, it was generally believed that in Africa south of the Sahara (southern Africa) the termi- nal Acheulian or “First Intermediate” gave rise to the Middle Stone Age (MSA) only 35 ky to 40 ky ago (Clark, 1970). This created the impression that Africa was something of a technological backwater during the later phases of Paleolithic prehistory. A further implication arising from this chronology was that relatively primitive, somewhat H. erectus-like hominids, such as Kabwe, persisted in Africa long after similar forms disappeared in Eurasia. All of this led Coon (1962) to infer that Africa south of the Sahara was one of the last regions of the world where modern humans arose from their archaic forerunners and led Clark (1970:120) to suggest a sudden replacement of the “Rhodesian physical type” by modern humans that had evolved in some other part of the world.

Chronometric dates from a number of sites now indicate that the MSA began over 100 ky ago (Deacon, 1989; Deacon and Geleijnse, 1988; Hay, 1987; Klein, 1983; Wendorf et al., 1975) and ended between 30 ky and 40 ky ago (Vogel and Beaumont, 1972). Morphological assessment of hominid remains associated (or possibly associated) with the MSA indicate that modern humans were established in Africa sometime during the MSA (Beaumont et al., 1978; Brauer, 1984a,b, 1989; Day and Stringer, 1982; Rightmire, 1979, 1984, 1986; Singer and Wymer, 1982; Smith, 1985, 1990; Stringer, 1989). Proponents of both the AES and RAE models believe that the modern anatomical pattern was established by approximately 100 ky ago. Furthermore, a number of researchers have argued that an unequivocal transition can be documented from archaic to modern humans in southern Africa (Brauer, 1983a, 1984a, 1990; Day and Stringer, 1982; Smith, 1985,1990; Stringer, 1988; Stringer and Andrews, 1988; Wolpoff, 1980). Obviously a detailed assessment of the post-erectus African fossil record, as well as the chronological framework in which this record is placed, is absolutely critical to evaluation of the competing models of modern human evolution.

The earliest modern Africans The argument for considerable antiquity of modern people in Africa is based

primarily on skeletal remains from Klasies River Main Site (KRMS), Border Cave, and cranium 1 from site KHS on the Omo Kibish Formation. Isolated teeth from the sites of Equus Cave (RSA), Die Kelders (RSA), and Mumba Rock Shelter (Tanzania) (Brauer and Mehlman, 1988; Grine and Klein, 1985) are often cited as further evidence for an early Late Pleistocene age for modern people in southern


Africa, since all of the teeth are morphologically and metrically indistinguishable from recent African dental samples (Brauer, 1989). However, these teeth would hardly build a convincing case were it not for the more anatomically informative specimens from Border Cave, KRMS, and Orno. Furthermore, recent assessment of Equus Cave suggests that the MSA associated with the hominid remains and artifacts are intrusive and that the hominid remains are probably latest Pleis- tocene or Holocene in age (Morris, 1990).

Cranial morphology can be evaluated in specimens from all three of the key sites. Both Orno 1 and Border Cave 1 are complete enough to judge overall cranial form as well as certain details of morphology. Both exhibit relatively vertical frontal squamae and high cranial vaults, which conform to the pattern found in most modern H. sapiens. Orno 1 exhibits both saggital and coronal vault contour profiles that approach closely those typical of most recent skeletal samples. Fi- nally, various multivariate analyses group both of these specimens with or near other early modern and recent African specimens and samples, rather than archaic ones (Brauer, 1983a, 1984a; Day and Stringer, 1982; Habgood, 1988; Rightmire, 19791, although Orno 1 does fall close to the more archaic Orno 2 cranium in the Day and Stringer’s Mahalanobis D2 analysis. Border Cave 1, Orno 1, and the KRMS 16425 frontal fragment all exhibit brow ridges that are reduced in overall size compared to archaic Africans, including Florisbad and Ngaloba, although they tend to be larger than is typical for recent Africans. Reduction occurs in both the anteroposterior projection and craniocaudal thickness of the brow ridges (see Smith and Ranyard [1980] for definitions of projection and thickness). Further- more, while the brow ridges do flatten craniocaudally lateral to the superciliary arches, there is less midorbital thinning and thus somewhat less distinct differentiation of the lateral and medial brow ridge segments (particularly in Border Cave 1-cf. Smith, 1990) than for most recent human samples.

Facial morphology is poorly known, but zygomatics from Orno 1, KRMS (166511, and Border Cave 1 are characterized by rather horizontal inferiolateral zygomaticoalveolar margins, a pattern typical of modern humans. The Border Cave 2 and 5 mandibles, as well as several from KRMS (e.g., 16424,41815), are clearly modern in form, but certain mandibles from KRMS appear slightly more primitive. KRMS mandibles 13400 and 21776, for example, exhibit rather vertical symphyses with little (21776) or virtually no (13400) mental eminence development.

Fragmentary postcranial remains are associated with Orno 1 and are described as exhibiting no features that fall out of the modern human range (Day, 1969; Howell, 1978). Kennedy (1984) has demonstrated that femoral morphology in Orno 1 clearly conforms to a modern pattern. The remainder of the Orno 1 postcrania have never been described in any detail. Also, a few postcranial fragments were recovered at Border Cave in deposits that slumped into a pit left by digging in the cave during the early 1940s. These specimens are robust and appear to be essentially modern in form, but their association with the purported MSA remains from Border Cave cannot be firmly established (Morris, 1990). Finally, the few postcranial fragments for KRMS described by Singer and Wymer (1982), especially the gracile clavicle (260761, also appear anatomically modern.

Based on the total morphological pattern of the Border Cave, Orno 1, and KRMS hominids, their designation as modern humans (Beaumont et al., 1978; Brauer, 1984a; Day, 1969; Day and Stringer, 1982; Rightmire, 1979, 1984, 1986; Singer and Wymer, 1982; Smith, 1985) is clearly justified. The somewhat primitive aspects of certain features in some specimens included in this sample, e.g., brow ridge development and mandibular symphyseal morphology, are to be expected in very early representatives of modern humans and help to establish their phylogenetic connection t o more archaic humans. The fact that multivariate statistical analysis of the Border Cave 1 cranium cannot unambiguously link it with any specific modern sample from southern Africa (Morris, 1990; Van Vark, 1986) indicates that these early modern Africans were not yet differentiated in the direction of any specific recent African population.


The African transition

If Africa is the (or even a) place of origin for modern humans, there should be evidence of an indigenous transition from very archaic H. sapiens-such as Kabwe, Bodo, Elandsfontein, Eyasi, and Ndutu-to the early modern sample discussed above. Stringer (1988; Stringer and Andrews, 1988) has recently listed the specimens he regards as documenting this transition. This African Transitional Group comprises the crania from Florisbad (RSA), Ngaloba (Tanzania), and Orno Kibish site DHS (Orno 2-Ethiopia1, all of which are also considered transitional by Brauer (1983a, 1984a, 1989). Stringer also includes the Jebel Irhoud hominids from Mo- rocco in the transitional sample. Rightmire (1978, 1984, 1986) sees a mosaic of features in the Florisbad, Orno, and Ngaloba specimens, some being more like the Kabwe group and others approaching the early modern forms. However, he adopts a more cautious view and stops short of suggesting that these unequivocally document an indigenous African transition.

Perhaps the most critical specimen to the transition issue is the Ngaloba (LH18) skull found at Laetoli in 1976 (Day et al., 1980). This specimen has a rather gracile face with a narrow nasal aperture, a horizontal inferiolateral zygomaticoalveolar margin, and a canine fossa. Vertical facial dimensions are also markedly reduced compared to the Kabwe group (Smith, 1985,1990). The cranial vault is rather low, with a frontal that places it more with the Kabwe group than with early modern hominids according to Brauer’s (1984a) multivariate analyses. The form of the parietals and aspects of occipital angulation and morphology are more modern-like (Rightmire, 1986). The supraorbital torus is well-developed and is intermediate in size between the Kabwe group and early modern humans (Smith, personal obser- vation).

The Florisbad cranium, newly reconstructed by Clarke (1985), exhibits similarities in the face to Ngaloba, specifically in the presence of a canine fossa, horizontal inferior zygomaticoalveolar margins, and reduced vertical facial dimensions compared to the Kabwe group. Frontal curvature is intermediate between the Kabwe group and recent Africans, and Brauer’s (1983a, 1984a) principal components analysis of frontal dimensions places Florisbad close to the modern African range. The supraorbital torus is similar to Ngaloba’s in form (Smith, 1990). Unfortunately, except for the anterior parietals, the remainder of the Florisbad skull is not preserved.

The Om0 2 cranial vault is demonstrably more primitive than Orno 1 and is very similar to archaic humans in occipital angulation, coronal vault contour, and low- ness of the cranial vault (Brauer, 1984a; Day, 1969; Rightmire, 1984). The archaic nature of Orno 2 is also reflected to some extent in the Mahalanobis D2 analysis of Day and Stringer (1982). In this analysis Orno 2 falls closest to Ngandong (from Indonesia), Kabwe and Orno 1 but not to other modern humans. Brauer and Leakey (1986) have also shown that the Eliye Springs (ES-11693) cranium is morphologically similar to the African Transitional Group, particularly by virtue of the obviously small face (particularly in vertical dimensions) and vault shape. Both Orno 2 and Eliye Springs are unfortunately surface finds, with no stratigraphic context.

Finally, the Jebel Irhoud cranial specimens from north Africa appear transitional in some analyses (Brauer, 1984a; Stringer, 1989). The Jebel Irhoud 1 cranial vault is basically archaic in form (Brauer, 1984a1, with a well-developed occipital chignon (Smith, 1985). The face, however, is relatively modern. It is short and exhibits a distinct canine fossa and modern-like zygomaticoalveolar region. Brau- er’s (1983a, 1984a) multivariate facial analysis places the Jebel Irhoud 1 face on the margin of the modern African range. The supraorbital torus is reduced compared to early archaic Africans but exhibits a pattern more like Europeans than other African specimens (Smith, 1990). The Irhoud 2 vault is higher than Irhoud 1 and falls within the modern range in Brauer’s (1984a) multivariate analysis of African frontal bones. It also exhibits an occipital chignon. Stringer (1989) has


described Jebel Irhoud 2 as perhaps the best example of a transitional specimen he has analyzed.

Based on the mosaic of archaic and modern morphometric features exhibited by these specimens, they do appear to represent a logical transitional series between the more archaic hominids of the Kabwe group and early modern Africans. How- ever, although Rightmire (1986) considers a transitional status for some of the African Transitional Group plausible, he makes one very important point. He notes that owing to the extensive geographic distribution and fragmentary nature of the specimens included in the African Transitional Group, it is difficult to “. . . identify any ‘clade features’ that may link archaic specimens from Broken Hill, Elandsfontein, and Lake Ndutu with later populations” (Rightmire, 1986218). In other words, the evolutionary continuity proposed for Africa is based on phenetic, not cladistic, evidence. What is present is a sequence documenting gracilization of the face, reduction of vault angulation, increase in vault height, and reduction of the brow ridges. But these are not shared derived (synapomorphic) characteristics unique to Africa and thus do not define an unambiguous African clade. Instead, these cranial vault changes characterize the emergence of modern humans throughout the Old World. The pattern of brow ridge evolution is one possible African clade feature (Smith, 1990), but it is not conclusive. Thus, it is possible to argue that the African transition to modern humans might not be the primary one, as both the AES and RAE assert, unless it can be clearly established that it predated evidence of transition in other regions. As Deacon (1989) so aptly puts it: “the plausibility of the interpretation of an African centre of origin for anatomically modern peoples . . . rests on the precision with which the archaeological evidence can be dated and with which the context of the finds can be established.”

Chronology and context Are there modern humans in southern Africa a t around 100 ky ago? Are the

fossils of the African Transitional Group older than 100 ky? Are the more archaic hominids (Kabwe/Bodo) older still? Such a temporal pattern is critical to both African origin models, so it is understandable that supporters of these models, while expressing cautious reservation about certain dates, regard the African chronological framework as generally reliable (Brauer, 1984a, 1989; Stringer, 1989; Stringer and Andrews, 1988). It is equally understandable that proponents of MRE have emphasized problems with context andlor precision of the date estimates for many key sites (Frayer, 1990; Smith, 1985, 1990; Wolpoff, 1989, 1990). The fact is that an objective assessment of the available chronological data would require answering each of the three questions raised above with a resounding “maybe!”

The early modern African sample with the fewest dating problems is that from Klasies River Main Site. This complex of overhangs and caverns was originally excavated in the late 1960s, and it was on the basis of these excavations that a claim for the presence of modern humans at 100 ky ago was made (Singer and Wymer, 1982). However, details of the stratigraphic correlation among the deposits in the various caves at KRMS (caves 1, lA, lB, lC, 2) and the dating of early MSA and Howiesonspoort levels has been questioned (Binford, 1984; Hendey and Volman, 1986; Parkington, 1989). Fortunately, recent excavations have greatly clarified the stratigraphic correlations and dating of the site (Deacon, 1989; Dea- con and Geleijnse, 1988; Deacon et al., 1986; Goede and Hitchman, 1987).

The pertinent deposits a t KRMS can be divided into three major stratigraphic units: the LBS, SAS, and Upper Members, each of which contains unequivocal evidence of human occupation (Deacon and Geleijnse, 1988). The LBS Member, which contains what Singer and Wymer refer to as the MSA 1 technological sub- stage, is the oldest and appears to date in excess of 100 ky ago. This is based on Shackelton’s (in Wymer and Singer, 1982) oxygen isotope correlation, amino acid dating (Bada and Deems, 1975), electron spin resonance (ESR) dating (Goede and Hitchman, 1987) and uranium disequilibrium dating. The latter suggests an age for the top of the LBS Member of between 98 ky and 100 ky ago (Deacon, 1989).


From the recent excavations at Klasies come two maxillary fragments that derive from LBS (Deacon and Geleijnse, 1988). Although undescribed as yet, these fragments have been characterized as robust (Deacon, 1989) and appear to represent a more archaic hominid form than the previously discovered KRMS hominids described by Singer and Wymer (1982).

The SAS Member contains Singer and Wymer’s MSA 2 and all of the MSA- associated hominids described by them in 1982. According to Deacon and Geleijnse (1988), this includes the robust but modern 41815 mandible from cave lB, which was previously associated with MSA 1. SAS is separated from LBS by an uncon- formity, and ESR analysis, but not lithic typology, of the base of SAS suggests a “significant temporal break between the two members” (Deacon and Geleijnse, 198853-9). The ESR analyses cannot provide precise age estimates for SAS (or LBS) owing to uranium enrichment in the zone of ground water percolation (Goede and Hitchman, 1987), so that it is not clear how much more recent than the LBS member the SAS member and its hominids are.

The Upper member, containing Howiesonspoort and MSA 3 assemblages, over- lies the SAS Member and is too old to be dated by conventional radiocarbon analysis (Deacon et al., 1986; Deacon and Geleijnse, 1988). Deacon uses amino acid dating by Bada and Deems (1975) and the presence of what is called the “Klasies River regression” (a lowered sea level that possibly correlates to oxygen isotope stage 4) to argue that the base of the Upper Member is >80 ky old. Thus, Deacon (1989) considers an age of between 80 ky and 100 ky to be reasonable for the SAS Member hominids.

Problems with date estimation for the Orno 1 and Border Cave hominids, the other potentially 100 ky old modern fossils, are somewhat greater. Orno 1 was found partially in situ at site KHS in a basal level in Member 1 of the Orno Kibish formation (Day, 1969; Day and Stringer, 1982). The specimen was associated with faunal remains including a primitive modern elephant (Loxodonta africana), an advanced archaic elephant (Elephas recki), and both white and black rhinoceros, a faunal association, which, according to Howell (1978:216), is “ . . . unconvincing of any remote antiquity.” Member 1 has an age estimation of approximately 130 ky ago on the basis of the uraniudthorium technique applied to mollusc shell (Butzer, 1969). Orno Kibish Member 3, which lies stratigraphically well above (ca. 60 m) the level of Orno 1, has been dated at >37 ky by radiocarbon dating of mollusc shell (Butzer, 1976). Butzer (personal communication) believes that Mem- bers 1,2, and 3 all lie within oxygen isotope stage 4 and thus are all older than ca. 75 ky. Citing this evidence, Brauer (1989) suggests that “ . . . an age of at least about 100,000 years seems well established for Orno 1.”

Four Border Cave specimens are claimed to derive from MSA levels, potentially in excess of 100 ky in age (Butzer et al., 1978). Of the four, only 2 were recovered in situ: the Border Cave 5 mandible and the Border Cave 3 infant skeleton. Both are described as fully modern in morphology and are associated with MSA levels dated at >49 ky BP by conventional radiocarbon (Beaumont et al., 1978; DeVilli- ers, 1973,1976). The Border Cave 1 partial cranium and 2 mandible were found by guano prospectors in the 1940s (Beaumont et al., 1978), and thus their context is uncertain. Chemical analyses, which appear to separate MSA and later bone samples a t Border Cave, support the claimed association of Border Cave 1 and 2 with MSA levels (Beaumont, 19801, but there are discrepancies among some of the analyses carried out by different researchers (Habgood, 1988). Butzer et al. (1978) claim ages for the levels containing the Border Cave hominids of between 90 ky and 115 ky BP based on correlations to oxygen isotope stages. However, Parking- ton (1989) has noted that correlations between an inland site like Border Cave and isotope changes in seawater involve making some very questionable assumptions. Furthermore, Klein (1983) has suggested that the Border Cave hominids may be intrusive into the MSA. He bases this on “. . . a strong contrast in state of pres- ervation between the human bones and animal bones that occur in the same levels. Unlike the animal bones, which are poorly preserved and highly fragmented as a


result of substantial postdepositional leaching and profile compaction, the human bones are relatively well-preserved and complete” (Klein, 1983: 34).

It is relatively safe to say that earlier archaic hominids like Kabwe, Elandsfon- tein, Ndutu, Bodo, and possibly Eyasi are late Middle or very early Late Pleis- tocene on the basis of archaeological, faunal, and geological evidence (Mehlman, 1987; Rightmire, 1984; Vrba, 1982) and that these are older than the African Transitional Group. However, most of the latter group are difficult to date pre- cisely. Omo 2 was a surface find, as was Eliye Springs; thus, both are undatable at present. Florisbad was recovered from the eye of an old spring, thus its exact stratigraphic position is uncertain. However, it was found at the stratigraphic level of Peat 1 (the lowest of 4 organic layers a t the site), associated with fauna and MSA artifacts. According to Clarke (19851, who has carried out new excavations at the site, preliminary uranium series dating indicates an age of >lo0 ky for Peat I and an age out of the range of conventional radiocarbon dating for the stratigraphically higher Peat 11. But, because the cranium was found in a debris-cone of the spring eye, it may be derived from a stratigraphically higher level than Peat I.

The dating of Ngaloba seems much more secure. The specimen was excavated in direct association with MSA artifacts and fauna. The geological bed yielding the cranium has been dated to between 90 ky and 150 ky BP based on geological correlation to nearby Olduvai Gorge (Leakey and Hay, 19821, to 129 k 4 ky (Th- 230) and 108 ? 30 (Pa-231) by uranium-thorium (Hay, 1987), and to 100-200 ky on the basis of isoleucine epimineralization (Bada, 1987). Also, the Jebel Irhoud hominids appear to date somewhere within oxygen isotope stage 5 (Hublin, personal communication), which would mean an age of between approximately 80 ky and 120 ky BP.

The African evidence-how we see it In our opinion, the difficulties in interpreting these African specimens lie more

in the realm of chronology than morphology. The Border Cave/KRMS/Omo 1 hominids conform morphologically to what one might expect of early modern Africans: a sample of hominids exhibiting a modern H. sapiens total morphological pattern, but with archaic-reminiscent features and not conclusively differentiated in the direction of any one recent African group. Unfortunately the dating and/or con- textual problems make i t essentially impossible to determine unequivocally an exact age for any of these specimens and thus for the appearance of modern humans in Africa. The KRMS hominids (and Border Cave as well, if one assumes they derive from MSA levels) lie out of the range of conventional radiocarbon dating, meaning they are minimally 40-50 ky old. The same applies to Omo 1 if one is willing to accept radiocarbon dates on mollusc shell (even though such dates are widely known to be extremely unreliable). The uranium/thorium date of 130 ky on mollusc shell, taken just above Omo 1, is also equivocal, since such dates on mollusc shell have been shown to have less than .5 probability of accuracy (see Hab- good, 1988). The only relatively secure upper limit (maximum) for the age of these early modern hominids is the approximately 100 ky BP date from KRMS. We assert that acceptance of ages significantly greater than 40-50 ky BP based on thickness of deposits between levels dated minimally to those ages and the strata containing the hominid remains is not judicious, given the well-known difficulty of estimating time from thickness of geological deposits. Thus, even though dates toward the older portion of the range may be conclusively demonstrated in the future, the best that can presently be said is that modern H. sapiens appeared in southern Africa sometime between 40 ky and 100 ky BP.

It seems to us that there are three possible explanations for the appearance of modern humans in Africa. The first is suggested by both “Out of Africa’’ (AES and RAE) models: there is an indigenous transition to modern humans in Africa, and they appear here first. This would be most clearly supported if the early modern Africans were toward the upper limit of the 40-100 ky range (and were of greater antiquity than early modern humans elsewhere), if the Transitional Group pre-


dates 100 ky, and if the more primitive archaic African hominids are older still. The second possibility is that modern morphology appears in Africa because of extra-Africa influence. Such a model can be extracted from the proposal by Van- dermeersch (1981b) to explain the ancestry of the Qafzeh hominids, but it is not clear if Vandermeersch would consider the emergence of modern western Asians as the primary appearance of modern peoples or what he sees as the connection between the western Asian and African early modern humans. An extra-African origin for early modern Africans has been mentioned as a possibility by Rightmire (1986), but not necessarily supported by him. Support for this interpretation would come through the demonstration that modern human morphology appeared outside Africa earlier than within Africa. Another possible indication would be evidence of considerable temporal overlap of the African Transitional Group and early modern humans (e.g., if Florisbad dates to ca. 40-50 ky and the early moderns to between ca. 60 ky and 80 ky). If this explanation is correct, it would mean that the transitional nature of the former group would be the result of parallelism or gene flow from outside Africa. The third possibility is that the African transition was part of a general, more widespread, pattern of evolution from archaic to modern humans (a multiregional phenomenon). We feel that this explanation would be supported if similar indications of transitional fossil samples could be identified in other regions of the Old World and if early modern humans could be shown to exhibit at least some regionally distinctive features or combinations of features in several different areas of the World.

While it may be argued that the first of these possibilities is the most defensible in light of existing evidence, we would simply caution that imprecision of dating makes it impossible to be as confident of the validity of either of the “Out of Africa” models as some of the proponents of these models seem to be. Establishment of a firmer chronology for Late Pleistocene African hominids is the key to determining which of these explanations is the most reasonable. Also, judgement regarding which explanation is the more parsimonious can only be made following a somewhat briefer discussion of the Late Pleistocene hominid fossil record from Eurasia.

WESTERN ASIA-ANOTHER EDEN?

Since 1980, considerable new information has also become available regarding Late Pleistocene human evolution in western Asia. Major analyses of two critical hominid samples, the Shanidar archaic humans (Trinkaus, 1983a) and the Qafzeh early modern forms (Vandermeersch, 1981a) have been published; and one very important new specimen, the archaic adult skeleton from Kebara, has been partially described (Rak and Arensburg, 1987). Additionally, both Vandermeersch and Trinkaus have offered interpretations of the course of hominid evolution in western Asia. Trinkaus (1983a, 1984) argues for a basically unilineal series of stages from an early archaic sample (mainly Zuttiyeh and the early group from Shanidar), through a late archaic sample (the late Shanidar group, Amud, and probably the Tabun C1 skeleton), and possibly to early modern western Asians (the SkhuUQafzeh group). Vandermeersch (1981b, 1989) derives modern humans directly from the Zuttiyeh hominid and argues that the other archaic hominids are relatively late-arriving Neandertal migrants from Europe. Finally, a series of new thermoluminescence (TL) and ESR dates suggest that modern humans have an antiquity of between 80 and 115 ky in this region, while archaic peoples cannot have disappeared before 60 ky BP (Bar-Yosef, 1990; Schwarz et al., 1988; Stringer et al., 1980; Valladas et al., 1987, 1988).

Various morphological analyses of the SkhulIQafzeh group establish that they are modern humans in both cranial and particularly postcranial form but also that certain primitive (archaic) features are present, particularly in the cranium (Mc- Cown and Keith, 1939; Simmons et al., in press; Stringer et al., 1989; Trinkaus 1983a,b, 1984; Stringer and Trinkaus, 1981; Vandermeersch, 1981a, 1989). These archaic-reminiscent features are, for the most part, clearly primitive retentions that could be derived from any archaic human forerunner (e.g., large brow ridges,


faces, and teeth); but there are some features (e.g., considerable alveolar prognathism and retromolar spaces in some specimens) that are more similar to Eur- asian archaic hominids than to the African Transitional Group.

The late archaic sample has a series of features that definitely identify them as Neandertals. These features are both cranial and postcranial (Trinkaus, 1984) but will not be detailed here, since the Neandertal morphological pattern is well known and has recently been reviewed in detail in several publications (Smith, 1984; Stringer, 1984; Trinkaus 1983a, 1986; Trinkaus and Smith, 1985). The early archaic sample including Zuttiyeh is similar to these Eurasian Neandertals in frontal morphology and other details but differs in having flat, albeit prognathic faces, rather than only midsagittally prognathic ones like the late archaic specimens (Trinkaus 1983a, 1984; Simmons et al., in press; Smith, 1985).

In addition to the interpretations offered by Trinkaus and Vandermeersch, proponents of the two “Out of Africa” models have suggested that the SkhulIQafzeh hominids are the result of African influence (Brauer 1984a, 1989; Stringer and Andrews, 1988). Stringer (1988), for example, sees them as part of a northern branch of the initial modern H . sapiens stock, which is ultimately of southern African origin. Each of these interpretations is possible, although Vander- meersch‘s rests entirely on the argument that Zuttiyeh is fundamentally different from later archaic hominids like Amud, Tabun, and the Shanidar late group and that these Neandertals are relatively late arrivals from Europe. He bases this primarily on Zuttiyeh’s facial flatness (Vandermeersch, 1981b, 1989). However, it has been demonstrated that this condition is not homologous to the true orthog- nathism of modern populations and is very similar to the early Shanidar group (Simmons et al., in press; Smith and Paquette, 1989; Trinkaus, 1984,1987). Thus, we regard it as very likely that Zuttiyeh is part of an early archaic group broadly ancestral to the later west Asian archaic humans (Tabun C1, Amud, Kebara, late Shanidar), although there are definite indications of European influence on the development of west Asian Neandertals both in a cladistic (Santa Luca, 1978) and phenetic sense.

As was the case for Africa, chronological problems pose considerable difficulty for interpreting the west Asian fossil record. The dates of the SkhulIQafzeh group have been debated for years (see Trinkaus, 19841, but new evidence in the form of TL and ESR dates support an early appearance for these early modern forms. TL dates for the Qafzeh hominids yield an age estimate of 92 ky BP (Valladas et al., 1988), and this has been supported by ESR age estimates of 96 ky or 115 ky BP, depending on whether a rapid or linear uptake of uranium is assumed (Schwarz et al., 1988). Very recent ESR dating of Skhul has produced age estimates of 81 ky or 98 ky BP, again depending on whether rapid or linear uptake of uranium is assumed (Stringer et al., 1989). When the ESR and TL dates are added to the geological and faunal evidence (Bar Yosef, 19901, the case for an approximately 90 ky age for the Qafzeh modern fossils is the strongest for an antiquity of this magnitude for early modern H. sapiens.

It merits some comment that the date of appearance for “typical” west Asian Neandertals is not clear. Bar-Yosef (1990) suggests ca. 80 ky BP, but the fact is that the ages for Shanidar, Amud, and Tabun C1 are simply not conclusively known. They could fall anywhere between ca. 50160 ky and ca. 100 ky BP. Plus, the possibility that Zuttiyeh is ancestral to the west Asian Neandertals pushes the age of this lineage back even further in this region.

In our opinion, it is not possible to be certain if the SkhulIQafzeh hominids evolved indigenously from west Asian archaic humans or as the result of some type of extraneous (presumably African) influence. Proponents of the “Out of Africa” models argue that the latter interpretation is more supported by available evidence, but this depends mainly on the chronology. As is true for Africa, there is little indication of regionally distinct features shared between west Asian archaic hominids and the SkhuUQafzeh group, but there is strong phenetic similarity between the SkhulIQafzeh and early modern African groups. Thus, in whichever


region modern morphology initially appeared or began to appear, it is likely to have had a strong influence on the other. We are inclined to think that Africa is the more likely area of origin for modern humans, mainly because we regard the African Transitional group as a more convincing source for the derivation of modern human form than the west Asian late archaic hominids. In the same vein, we believe that the gracilization of the face in Africa might have served as a catalyst for the facial reduction demonstrated by Trinkaus (1984) for western Asia. How- ever, features like the pattern of prognathism and retromolar space retention in some of the SkhulIQafzeh sample may reflect a considerable Eurasian contribution to early modern west Asians.

If, however, the Qafzeh hominids turn out to have greater antiquity than the early modern Africans and/or the African Transitional group turns out not to be as old as is presently thought, our stand on this “chicken or egg” question would have to be reconsidered. At any rate, we believe that an earlier appearance of modern humans in western Asia than Africa is a possibility that cannot be cursorily dismissed and that a certain degree of continuity between both groups of archaic and early modern hominid samples in west Asia is a distinct possibility.

THE PATTERN AT THE PERIPHERIES EUROPE, EAST ASIA, AND AUSTRALASIA

Both “Out of Africa” models predict that modern humans should appear relatively late in these areas and require that there be no evidence of transition outside Africa. Thus, nothing comparable to the African Transitional Group should be present in any other region of the Old World (Brauer, 1984a,b, 1989; Stringer, 1989, 1990; Stringer and Andrews, 1988). While Brauer’s AES model allows for minimal continuity in early modern Eurasians, the RAE model apparently does not. The Multiregional model not only predicts the existence of transitional samples in regions other than Africa, but certain adherents (principally Wolpoff, 1989, 1990) have suggested that non-adaptive, regionally specific “racial” features would have greater antiquity in these “peripheries” than in the “central” areas. This would be the result of mechanisms promoting regional differentiation during the Homo erectus radiation and the fact that gene flow is predominantly unidirectional into peripheral populations. Discussion of these “peripheral” regions will revolve around identifying the date for the earliest modern humans and the evidence for transitional groups and clade (regionally distinctive) features that cut across archaiclmodern human boundaries.

Europe-the ultimate periphery The European Late Pleistocene fossil record is still the best known and under-

stood in the Old World. It is clear that Neandertals evolved from an earlier, more H. erectus-like group (consisting of specimens such as Petralona, Arago, Swans- combe, and others) that become progressively more Neandertal-like during the late Middle Pleistocene (Hublin, 1988; Trinkaus, 1986; Vandermeersch, 1985; Wol- poff, 1980). The erectus-like group is broadly similar to the African Kabwe group, but the pattern of morphological change diverges between Africa and Europelwest Asia after this point. Rather than the facial gracilization and size reduction characteristics of the African Transitional Group, European Neandertal faces remain large and develop a series of specializations, including slightly increased amounts of mid-sagittal facial prognathism (especially at nasion) and a more sagittally oriented infraorbital plate (Rak, 1986; Trinkaus, 1983a, 1987; Smith, 1983; Smith and Paquette, 1989). Brow ridges, however, were reduced, as was the degree of lateral facial prognathism. Neandertal postcrania are characterized by a general continuation of the basic pattern seen in H. erectus, albeit with some reduction in cortical shaft thicknesses and other features (Kennedy, 1990; Trinkaus, 1983a, 1986). There appear also to be some specializations in Neandertal postcrania. These include low brachial and crural indices, elongated and thinned superior pubic rami, and other features relating to muscular hypertrophy (Rosenberg, 1988; Stringer and Trinkaus, 1981; Trinkaus, 1983a,b, 1989). Neandertal features are


evident very early in ontogeny (Heim, 1982; Minugh-Purvis, 1988; Tompkins and Trinkaus, 1987), and there are indications that Neandertals reached dental ma- turity faster (Wolpoff, 1979) and tended to die younger than modern humans (Trinkaus and Thompson, 1987).

Early modern humans in Europe conform generally to the cranial pattern discussed for early modern western Asians and Africans. Postcranially, they are rather different from Neandertals, as Trinkaus (1983a,b, 1986, 1989) and others (e.g.! Kennedy, 1990) have shown. Generally, they lack the thick cortices and specializations characteristic of Neandertals, but there are still some general indications of continuity in terms of joint size (Wolpoff, 1989) and femoral form (Kidder, 1988). If the dating is correct, modern humans appear very late in western Europe-maximally around 30 ky at the site of Cro-Magnon (Stringer et al., 1984). The Neandertal skeleton from St. Cesaire shows that typical Neandertals existed in western Europe until about 33-35 ky BP (Lev6que and Vandermeersch, 1981; Leroyer and Leroi-Gourhan, 1983). Furthermore, there is little evidence of evolutionary trends in the modern human direction among the west European Nean- dertals (Stringer et al., 1984; Smith, 1985), except for dental reduction (Wolpoff, 1980). In central Europe, modern humans appear around 34-36 ky BP, a t sites like Hahnofersand (Germany), Velika PeCina (Yugoslavia), and MladeE (Czechoslova- kia) (Brauer, 1980; Frayer, 1986; Smith, 1982,1984). However, in central Europe there are possible indications of diachronic trends within the Neandertals, in the direction of the modern European condition (Smith, 1982, 1984, 1990; Smith and Trinkaus, in press; Wolpoff, 1980, 1989; Wolpoff et al., 1981). In southern Europe (Greece and Balkans), possibly early modern humans have been reported from Bacho Kiro Cave (Bulgaria) a t a date of >43 ky BP. However, upon analysis (Glen and Kaczanowski, 1982) these specimens (essentially teeth) fall closer to Nean- dertals. Also, a fragmentary cranial vault and partial postcranium from the island of Crete has recently been reported. This specimen, found in 1893, has been dated by the ProtoactiniudUranium method to 51 5 12 ky BP (Facchini and Giusberti, 1990). The specimen is described as modern H. sapiens but with retention of archaic features. Details of frontal curvature and other vault features suggest strong similarity to early modern central Europeans (Facchini and Giusberti, 1990).

The most convincing evidence of continuity in Europe is found in central Europe. Here a series of late Neandertals, consistin of remains from Vindija Cave (Yugo-

ing to approximately 38-45 ky BP, exhibit a pattern of facial reduction similar to the African Transitional Group. For example, facial dimensions are reduced, canine fossae are present, mandibular symphyses tend to be more vertical, and brow ridges are reduced (Smith, 1982, 1984, 1990; Wolpoff et al., 1981). The meager postcranial sample reveals a pattern of change only in the scapular glenoid fossa proportions (Smith and Trinkaus, in press). Furthermore, the earliest modern specimens from central Europe exhibit a number of characteristics that have been interpreted as reflecting some degree of European Neandertal ancestry. These include general vault shape (particularly in males), the high frequency of occipital bunning (although reduced from the Neandertal condition), the size and form of the brow ridges, and a series of other features (Frayer, 1986, 1990; Smith, 1984, 1990; Smith and Ranyard, 1980; Smith et al., 1989; Smith and Trinkaus, in press; Wolpoff, 1980, 1989).

Although many features do not change abruptly at the Neandertallmodern Eu- ropean boundary, others do seem to exhibit a more marked change, probably reflecting increased effects of incoming gene flow (Trinkaus and Smith, 1985). Pro- ponents of replacement models argue that this change is due to the influx of modern populations and that the indications of a transitional group in Europe are equivocal. They suggest that the observed changes from early to late Neandertals are either parallelisms or are related to an overabundance of females and juveniles in the Vindija sample (e.g., Brauer, 1990). The so-called Neandertal reminiscent features of early modern samples are explained as either primitive retentions that

slavia), K6lna Cave (Czechoslovakia), and 73 ipka Cave (Czechoslovakia) and dat-


do not uniquely connect early Europeans with Neandertals (e.g., large brow ridges) or as homoplasies that do not reflect ancestry (e.g., occipital bunning) (Stringer and Andrews, 1988). Finally, it is argued that Neandertals and modern humans overlapped in western (Stringer et al., 1984) and possibly central (Allsworth-Jones, 1986) Europe.

In our opinion, while the changes that characterize the origin of modern Euro- peans are not to be taken lightly, neither are the indications of continuity. The trends in the Neandertal sample are real. It is possible that they represent parallelisms, but they are not the result of an overabundance of smaller or subadult individuals at Vindija (Smith, 1984). Furthermore, it is the pattern of brow ridge reduction not just reduction alone that is a potential reflection of European continuity. The European pattern involves a progressive mid-orbital diminution that does not appear to characterize the African clade (Smith, 1990), although the pattern could possibly be derived from elsewhere or result from drift or some other microevolutionary process. For example, this pattern could possibly have derived from North Africa, where the Jebel Irhoud specimens exhibit a similar brow ridge pattern, occipital buns, and a Neandertal-like humerus (Hublin et al., 1987), but it is equally likely that the presence of these features in North Africa is the result of gene flow from Europe into this region. Finally, it is not clear that Neandertals and modern humans overlapped. In western Europe this is based on evidence of overlap between Aurignacian and Chatelperronian archaeological components, but the assumption that this means interstratification of hominid types is just that: an assumption. In central Europe, the evidence for overlap is that the Sipka Neandertal and Mladei: early modern hominids come from the same interstadial. The Qipka mandible, however, is not that different from early modern ones and would not be out of place in such a sample (Smith, 1982, 1984).

Realistically, it is difficult to posit an absence of significant extra-European influence in the emergence of modern Europeans, particularly given the apparent southeast to northwest pattern of appearance of modern human anatomical features in Europe. But the indications of continuity across this transition and the existence of a transitional morphological pattern (albeit probably later in time than in Africa) suggest to us that population replacement is not the most likely explanation for this phenomenon. Certainly the evidence that favors replacement-with or without hybridization-as the mechanism responsible for modern European origins is equivocal, and in our opinion much less convincing, than proponents of such models assert. On the other hand, the indications of continuity are not strong enough to suggest that extra-European influences played only a minor role in the origin of modern Europeans.

East Asia East Asia is taken here to encompass the regions north of tropical southeast Asia

and essentially consists of the present countries of China, Japan, and Korea. There is a relatively extensive series of Homo erectus remains from this region (Pope and Cronin, 1984; Wu and Dong, 19851, but relatively little in the way of Homo sapiens remains (Wu and Wu, 1985). Archaic H . sapiens is represented primarily by the cranial remains from Dali and Maba and the partial skeleton from Jinniu Shan or Yingkou, all from China. Archaic H . sapiens (or evolved H . erectus) has also been reported from south central Asia (India) in the form of a calvarium from Narmada (De Lumley and Sonakia, 1985). Dates for all of these specimens are approximate. Uranium series dates for Jinniu Shan suggest an age of 100-200 ky (Pope, 19881, and the others are associated with fauna suggesting a late Middle or early Late Pleistocene age. These specimens all have relatively low cranial vaults, prominent supraorbital tori, and other features commensurate with their designation as post- H . erectus, pre-modern human specimens (Pope, 1988; Smith, 1985; Wolpoff, 1985; Wolpoff et al., 1984; Wu and Wu, 1985). The faces of Dali and Jinniu Shan exhibit rather short vertical dimensions, canine fossae, horizontal inferior zygomaticoalveolar margins, and (apparently) only moderate prognathism (Habgood, 1990;


Pope, 1988; Wu and Wu, 1985). The yet-undescribed Jinniu Shan specimen also preserves a partial postcranial skeleton that is suggested to be essentially modern in form.

The date of appearance for early modern humans in East Asia is very problematic. The earliest specimens exhibiting a modern morphological pattern are from the Upper Cave at Zhoukoudian, Ziyang, Liu Jiang, and a few others (Wu and Zhang, 1985). The Upper Cave specimens are dated by radiocarbon to between 10 ky and 18 ky BP, Ziyang may be approximately 7 ky old, and the others are geologically dated to “Late Pleistocene” (Wu and Zhang, 1985; Zhou et al., 1982). Since there is a large temporal gap between the latest archaic and earlier modern forms in East Asia, it is not clear when modern humans first emerged in East Asia.

The other major question regarding these hominids deals with the issue of regional continuity. It has been argued that H . erectus exhibits considerable stasis and not much indication of a temporal trend toward the archaic H . sapiens condition in Asia and Africa (Rightmire, 1985). Others have noted a number of features that seem to establish such a diachronic trend (Wolpoff, 1985; Wolpoff et al., 1984; Wu and Dong, 1985). Pope (1988:61), for example, lists increasing cranial capaci- ties, higher frontals, reduced post-orbital constriction, and reduced supraorbital torii as differentiating late from early Asian H . erectus. When one considers the similarity of Dali to late H . erectus, it is difficult to deny the existence of a sequence with strong phenetic indications of continuity between H . erectus and archaic H. sapiens in China.

The existence of continuity between the archaic and early modern H. sapiens is less convincing. A number of researchers, from Weidenreich to Wolpoff, have suggested that a number of features indicate continuity in this region across the archaidmodern human boundary (reviewed in Wolpoff et al., 1984; and Habgood, 1990). However, detailed assessment suggests that most of these are not unique clade features for East Asia as they are commonly found in modern samples outside this region, are not characteristic of all Mongoloid samples, and/or are con- sistently present in archaic H. sapiens and/or H . erectus throughout the Old World (Brauer, 1989; Habgood, 1990; Stringer and Andrews, 1988). On the other hand, a series of features reflecting facial flatness seem to be rarely found outside East Asia and not generally in combination on the same specimens outside of this geographic area (Habgood, 1990; Wolpoff, 1985; Wolpoff et al., 1984). According to Habgood (19901, the features, which document some degree of “regional continuity” in East Asia, are a non-depressed nasion, more perpendicularly oriented nasal bones, frontomaxillary sutures on almost the same level, and an angular rather than rounded junction of the zygomatic bone and zygomatic process of the maxilla. Thus, assuming Habgood’s analysis is correct, it seems improbable that the appearance of modern humans in East Asia resulted from a complete replacement. Both Brauer (1984b) and Howells (1983) believe that physical migration of peoples of modern form into East Asia played the dominant role in the establishment of modern human morphology in this region, although they both accept the possibility of assimilation of some archaic East Asian elements. However, the extent to which actual migration of peoples in East Asia occurred cannot be determined at the present time owing primarily to the inadequacies of the preserved fossil record.

Australasia As defined here, Australasia consists of the tropical regions of southeast Asia

and the islands of the Sunda and Sahul shelves (Australia and New Guinea). There is a certain arbitrariness to this definition, particularly since Sahul and Sunda were not joined by land connections at any time pertinent to discussions of modern humans. Nevertheless there has long been a tradition of recognizing a very close relationship between the late Pleistocene fossil hominid samples from these regions (cf. Wolpoff et al., 1984).

As he did for China, Weidenreich suggested quite early that a lineal connection from H . erectus to modern people existed in this region. His perspective, reviewed


in Habgood (1989) and Wolpoff (1985, 1990; Wolpoff et al., 19841, was that distinct similarities connected Indonesian H . erectus, through the Solo hominids, to Aus- tralian aborigines. Recent advocates of the MRE model have expanded on Weiden- reich‘s argument, particularly on the basis of Wolpoff‘s reconstruction of the San- giran 17 face, and have detailed a number of features that they claim indicate an Australasian clade (Thorne and Wolpoff, 1981; Wolpoff, 1985, 1986, 1989, 1990; Wolpoff et al., 1984). Opponents of regional continuity have responded that the similarities used to define this “clade” are not suitable for such purposes since they are generally primitive retentions or have distributions in archaic and modern groups that negate their use as Australasian clade markers (Brauer, 1989; Green and Groves, 1989; Stringer and Andrews, 1988). Two other problems for the MRE model are 1) some of the potentially earliest modern specimens in this region (Mungo 1 and 3 from Australia and the “deep skull” of Niah) do not fit easily into an Australasian clade characterized by craniofacial robusticity, and 2) there is little in the way of possible archaic H . sapiens material bridging the gap between H . erectus and modern human skeletal remains.

The main hominid sample that qualifies as possibly archaic H . sapiens is from Ngandong (12 partial crania and 2 tibiae) in Indonesia. The actual date of this series continues to be problematic, but these hominids do appear to be stratigraphically younger than the Kabuh-aged Indonesian H. erectus (Bartstra et al., 1988). In the most detailed study of these hominids available, Santa Luca (1980) con- cludes that, despite increases in cranial capacity and concomitant changes in vault form, the Ngandong hominids are best considered late Homo erectus. Stringer (1984) agrees with Santa Luca but sums up the Ngandong dilemma quite well by stating that ‘l. . . some individuals in the sample certainly differ from the typical H . erectus S . S . morphology in ways approximating that of H . sapiens s.1.” (Stringer 1984: 139). The Ngandong crania do share some striking similarities with earlier Indonesian H . erectus, but the issue of whether they are transitional between H . erectus and modern H . sapiens or the terminus of a southeast Asian H . erectus lineage rests with the ability to demonstrate the likelihood of a connection between Ngandong and modern Australasians.

In the most comprehensive assessment of this issue to date, Habgood (1988, 1989) has evaluated the features suggested by Wolpoff and his coworkers to demonstrate an Australasian clade extending back to Indonesia H. erectus. Habgood’s analysis demonstrates that many of these features are not suitable as unequivocal clade features, because they are generally characteristic of H . erectus and archaic H. sapiens in other regions. However, Habgood does recognize that there is a group of features (some of which are symplesiomorphies), which, in combination, may document existence of a “morphological clade” in Australasia. These include a long and sagittally flat frontal bone with a posterior position of minimum frontal breadth (in undeformed crania), very prognathic faces, and malars with everted lower margins and prominent zygomaxillary tuberosities. Wolpoff and coworkers also emphasize that it is the combination of features, rather than individual features that characterizes the Australasian pattern (Thorne and Wolpoff, 1981; Wol- poff et al., 1984). It is important to note, however, that the only H . erectuslarchaic H. sapiens specimen preserving a face (necessary to determine degree of prognathism and zygomatic form) is Sangiran 17. Furthermore, the Ngandong hominids can only be directly integrated into this scheme on the basis of features on the frontal bone.

The final issue concerns the dating of early modern humans in this region. On Sunda, the Niah skull (Borneo) and the fragmentary Tabon hominids (Philippines) are generally considered the earliest modern representatives. The former is considered to be about 40 ky old based on radiocarbon dates (Kennedy, 1979). How- ever, the stratigraphic interpretation of Niah Cave has been questioned, and it is possible that the relatively gracile, probably adolescent cranium is intrusive into the 40 ky old level (Bellwood, 1985). The specimen seems to have affinities with both gracile recent Australasian and mainland Asian samples (Brothwell, 1960;


Habgood, 1989). The Tabon specimens date to approximately 23 ky BP (Oakley et al., 1975), are relatively gracile, and have been interpreted as also having similarities to both modern Australian and mainland Asian samples (Wolpoff et al., 1984).

On Sahul, the relatively gracile Mungo 1 and 3 remains date to between 24 ky and 30 ky BP, and the much more robust Lake Garnpung (Wallandra Lakes Hom- inid 50) specimen appears to date between ca. 25 ky and 45 ky BP, but could be even older (see dating review in Habgood, 1988). The latter fossil is critical as it is the most archaic (Ngandong-like) specimen yet to be found in Australia and is suggested to be an excellent transitional form between Ngandong and recent Aus- tralians (Thorne, personal communication; Wolpoff, 1985, 1990; Wolpoff et al., 1984; Smith, 1985). Unfortunately, the specimen is not yet described in detail. The Mungo remains, which are on the gracile end of the modern Australian range of variation, may either reflect Asian influences in Australia (as a part of Thorne’s dual origin hypothesis) or be the result of demographic and genetic factors acting on a single founding lineage during the Late Pleistocene (Habgood, 1988; 1989; Thorne, 1976; Wolpoff et al., 1984). It could also be argued that the Mungo hominids (along with Niah and Tabon) represent the earliest migrants into the area and that the robust features ascribed to later fossil Australians (e.g., Kow Swamp, Cohuna, etc.) were the result of more recent microevolutionary adaptive factors. This would, of course, offer support for a replacement explanation of the emergence of modern people in this region. If the Lake Garnpung cranium (WLH 50) proved to have an antiquity of >30 ky, this latter possibility could be dismissed. The recently reported ESR date of 25-30 ky for this specimen (see Stringer, 1990) is a minimum date according to Wolpoff (personal communication) and thus does not refute a 30 + ky age for the specimen.

Obviously, the situation in Australasia is just as complicated as in the other regions discussed. We believe that the evidence of at least some continuity in this region makes replacement/speciation models like RAE highly unlikely. We are not convinced, however, that existing evidence allows a confident choice between an essentially replacement-oriented model (like AES) or the more continuity-oriented ME model as an explanation of Late Pleistocene hominid evolutionary patterns in Australasia. There is nothing that clearly demonstrates the migration of fully modern populations into greater Australasia, but the evidence in terms of fossil remains for this period is so limited that it certainly cannot be categorically denied that such a migration occurred. On the other hand, if it turns out that the gracile early modern hominids in this region do represent the initial modern group in Australasia, a replacement with a hybridization model, such as Brauers AES model, would gain added support. Such a model would, as Brauer (1989) asserts, explain the Ngandong-like aspects of some more recent Australians. But, should WLH 50 date earlier than the Mungo specimens, it would suggest that the transition to modern humans was well advanced prior to the appearance of these gracile modern humans.

THE GENETIC FACTOR

In the years since Howells’ (1976) influential assessment of modern human origins, data on genetic variability of extant human populations have become increasingly important to interpretations of the evolutionary history of modern people. Summarizing earlier work by Edwards and Cavalli-Sforza, Howells suggested that modern human genetic data support a single-origidradiation model, but disagreed with the view that genetic differentiation of modern populations results entirely from stochastic factors. Cavalli-Sforza’s (1974) blood group data showed a smaller genetic distance between Africans and Europeans than between Africans and Asians, but more comprehensive studies utilizing a total of 85 blood groups and protein systems revealed a consistent split between Africans and Eur- asians (Nei, 1985; Nei and Roychoudhury, 1982). From these data, divergence times were calculated. The African-Eurasian split is estimated to have occurred


approximately 110 2 34 ky ago, while the EuropeanlAsian divergence is set a t 41 * 15 ky ago (Nei, 1985). Recent work by Cavalli-Sforza and his coworkers also documents an Africanlnon-African split (Cavalli-Sforza et al., 1988).

Genetic evidence for a recent African origin of all modern human populations has been offered based on studies of both mtDNA by Cann and her colleagues (Cann, 1987,1988; Cann et al., 1987; Stoneking and Cann, 1989) and nuclear DNA by Wainscoat’s group (Wainscoat et al., 1986; Hill and Wainscoat, 1986). Cann et al. (1987) utilized restriction endonucleases to map approximately 9% of the mtDNA of 147 humans from Africa, Asia, New Guinea, and Europe. They found that mtDNA diversity is greater in peoples of African ancestry than in those of any other region. Furthermore, the phylogenetic tree constructed from the 133 recognized mtDNA types (utilizing maximum parsimony) by Cann and her coworkers reveals two major clusters: one entirely African and one mixed Africadnon-Af- rican. On the basis of these results, it was concluded that African populations have been diversifying longer than those from other regions and thus that Africa was the homeland of the modern human lineage.

Cann et al. (1987) also argue that changes in mtDNA are selectively neutral and consequently accumulate at a constant rate over time. Based on estimates of when humans reached Australia, New Guinea, and the New World, they calculate that mtDNA of human populations diverge at a rate of 2-4% per million years. Utiliz- ing this rate, Cann et al. estimated that the common ancestor of all surviving mtDNA types existed between 140 ky and 290 ky BP and that the ancestral stock of all Eurasians diverged from an African one between 90 ky and 180 ky BP. Since mtDNA diversity is very low in Eurasian populations, proponents of a constant rate of change for mtDNA suggest that assimilation of mtDNA from archaic human populations in Eurasia (i.e., hybridization between the modern stock from Africa and Eurasian archaics) was either minimal (Cann, 1988) or essentially nil (Stoneking and Cann, 1989). The reason for denying the possibility of assimilation of archaic elements into the Eurasian gene pool is clearly stated by Cann et al. (1987:35):

If there was hybridization between the resident archaic forms in Asia and anatomically modern forms emerging from Africa, we should expect to find extremely divergent types of mtDNA in present-day Asians, more divergent than any mtDNA found in Africa.

Since (according to Cann et al.) there is no evidence for these “extremely divergent types,” either there was no archaic contribution to the modern gene pool or there were additional factors (other than accumulation of neutral mutations) influencing mtDNA variability.

Wainscoat and his coworkers (Hill and Wainscoat, 1986; Wainscoat et al., 1986) have also argued for an African origin of modern humans based on analysis of the p-globin gene cluster. The distributional patterning of the five most useful haplotypes in this cluster is interpreted as revealing a basic Africadnon-African split. The differences between the two divisions are explained by an African origin, followed by passage through a “bottleneck, for the non-African stock. The appearance of two otherwise uniquely African haplotypes in Melanesia is explained as an example of homoplasy.

Supporters of both “Out of Africa” models have argued that the genetic evidence offers virtually complete support of a recent African origin of modern people (Brauer, 1989; Gould, 1987; Stringer and Andrews, 1988). It is important to emphasize, however, that there is not unanimity among geneticists regarding how these genetic results should be interpreted or even how reliable they are, particularly as clocks. The use of genetic data from extant peoples to construct phylogenetic trees is based upon a number of assumptions, which are almost always violated by real populations (Harpending, 1974). For example, such models not only assume that population differentiation is primarily the result of neutral ev-

56 YEARBOOK OF PHYSlCAL ANTHROPOLOGY [Vol. 32, 1989

olution, but also that the groups are isolated from one another following divergence. Such a methodology is simply not well suited for work on modern people, who have always been highly mobile and continue to exchange genes throughout their range. As Cann (1988:134-135) puts it, such methods ‘l. . . imply the cessation of gene flow, and those measures when used to estimate the age of separation of human races have little biological reality for our species.” Similar conclusions are also reached by Weiss and Maruyama (1976) and Wolpoff (1989).

Specifically regarding the P-globin studies, Giles and Ambrose (1986) note that these data are equally compatible with a Eurasian origin of modern humans followed by a migration to Africa. Also Wolpoff (1989) points out that the different ways in which the data from Africa and Europe were grouped may have the effect of making the Africans seem more distinct from the rest of the world than may be the case. While phylogenetic trees are clearly useful as heuristic devices, we tend to agree with Morton (cited in Harpending, 1974:237):

Phylogenetic trees are like flower arrangements; it is enough that they are pretty, without asking that they be meaningful as well.

There are several researchers who argue that the mtDNA results do not demonstrate a recent African origin for all modern humans. Using somewhat different analyses than Cann and her coworkers, both Asian and Caucasoid populations have been suggested to be possibly more suitable for the ancestral stock of modern human mtDNA than Africa (Johnson et al., 1983; Excoffier and Langaney, 1989, respectively). However Excoffier and Langaney (1989) state that a hypothesis suggesting primitive modern humans were Caucasoids or that they actually appeared first in the Caucasoid geographic region is not testable, though it is supported by data on Rhesus, Gm, and HLA genetic systems. Furthermore, Excoffier and coworkers (Excoffier et al., 1987; Excoffier and Langaney, 1989) present evidence that selection may have played a role in mtDNA evolution and that the “genea- logical tree” on which the African origin model of mtDNA is based is ll. . . biased by topological errors.”

Both Spuhler (1988) and Excoffier and Langaney (1989) provide data indicating that there is significant regional variation and considerable time depth to existing human mtDNA differences. Spuhler argues that many mtDNA types are unique to and very old in regions outside Africa and that their ancestry can be extended to time ranges prior to the estimates of mtDNA radiation into Eurasia by Cann et al. Excoffier and Langaney (1989232) note that most of the very differentiated mtDNA types in African populations ll. . . have appeared only recently.”

The low mtDNA variability of humans may be explained by the process of stochastic lineage extinction (Avise et al., 1984: see also Spuhler, 1988; Wolpoff, 1989, 1990) rather than as a result of all recent humans having evolved from “. . . a small mitochondrially monomorphic population . . .” (Brown, 1980:3609). Because of its haploid mode of inheritance, mtDNA lineages are very likely to become extinct simply as a result of random chance. In each generation, by chance, some females will produce no daughters and will therefore contribute no mtDNA to the next generation. If a generation of females bearing a mtDNA type all have no daughters that lineage will become extinct. Under most conditions the stochastic extinction of mitochondria1 lineages will be rapid, and the chances of a lineage remaining extant indefinitely are quite small (Avise et al., 1984).

With respect to its transmission, mtDNA is analogous to surnames. In many societies surnames are transmitted from the father to all of his offspring, and surnames, like mtDNA, behave as an asexually transmitted trait in a sexually reproducing species (Yasuda et al., 1974). Lotka (1931a,b) studied the extinction of surnames from a theoretical perspective and demonstrated that the probability of a male lineage remaining extant is small. When the extinction of surnames in a real population was studied it was found that the observed frequency of surname extinction is very close to the expected frequency (Yasuda et al., 1974). Yasuda et


al. (1974) studied the extinction of surnames in a population over a period of 300 years, approximately 10 generations, and found that over that period of time about 80% of the male lineages initiated by a single male became extinct.

The extinction of mitochondrial lineages, like the extinction of surnames, will usually be rapid. The rate at which lineages will become extinct will vary, but under certain demographic conditions it will occur very quickly. Avise et al. (1984) have examined the parameters that affect the rate of stochastic lineage extinction. They found that in a stable population initiated by n females it is highly probable that within 4n generations all of the descendants will trace their ancestries to a single female founder. The probability that two or more lineages will remain extant is high only when the number of generations since the population was founded is less than one-half the number of founding females.

Under certain demographic conditions stochastic lineage extinction will be ac- celerated or retarded. One factor that influences lineage survival is the variance in the number of females born; increased variance decreases lineage survival. For example, in a stable population founded by 15,000 unrelated females, with a variance in the number of daughters born of 1.0, there is a 50% probability that within 18,000 generations the mitochondria of all individuals will be derived from a single female. In a stable population of 45,000 unrelated females with a variance of 3.0 there is also a 50% probability that within 18,000 generations all individuals will have mtDNA derived from a single female (Avise et al., 1984).

One other factor that Avise et al. identified as having an important effect upon the rate of lineage extinction is the rate of growth of a population. In a population that is decreasing in size lineage extinction occurs very rapidly, and the probability that two or more lineages will remain extant quickly becomes zero, even when there is a large founding population. On the other hand, in a population that is expanding in size the probability that two or more lineages will survive will be large regardless of the size of the founding population.

Cann et al. (1987) argued that the low diversity found in the mtDNA of modern humans could not be the result of lineage extinction because the probability of lineages becoming extinct is very small in expanding populations. Although the human population has expanded dramatically since the Pleistocene, prior to that time human populations have usually been growing very slowly, or not a t all, for long periods of time (Weiss, 1984). In fact, Weiss (1984) has shown that if the human population had not been relatively stationary in growth, or nearly so, for most of its history the present population of the earth would be much greater, by a factor of approximately 30, than its present size of 4.5 billion. Thus, the apparent growth pattern of Pleistocene human populations would not prevent lineage extinction from reducing mtDNA variability in humans.

The mtDNA data do not reveal anything about the contributions made to modern gene pools by males and females whose mtDNA has not survived to the present; because of its mode of inheritance, and the consequent stochastic extinction of lineages, the great majority of our ancestors whose nuclear genes we carry have contributed no mtDNA to the modern mitochondrial gene pool. Every individual living today carries the nuclear DNA of a large number of ancestors, and represents a combination of nuclear genes contributed by 2" ancestors who lived n generations ago. Yet that individual carries the mtDNA of only one of those ancestors in any generation. It does not follow that the individual is descended only from that one ancestor.

Latorre et al. (1986) studied mtDNA variation in Drosophila subobscura, a species of fly that is widely distributed in the Old World and that has recently colo- nized the New World. They note that:

Today's world population of D. subobscura consists of many millions of individuals. It might very well be the case that, a few hundred thousand years hence, all D. subobscura flies have mtDNAs derived from morph I. This would not mean that the mtDNA of the descendants derives from only one D. subobscura cur-


rently living-morph I is found in 44% of the living population. More impor- tantly, the individuals living in that remote generation would count among their ancestors not only those females from which they inherited their mitochondria but also innumerable other females and males from which they inherited their nuclear hereditary material (Latorre et al., 198623652-8653).

A similar situation must also apply for humans. Even if all living humans share mtDNA that can be traced back to a single morph, perhaps even one that origi- nated in Africa, they also have nuclear DNA from countless other ancestors many of which need not have been African.

Finally there is the issue of the rate of evolution in mtDNA. The 2-4% per million year rate utilized by Cann et al. (1987) has been questioned. The estimate of 2% is based on a divergence time of 5 mya between Pan and Homo; but if divergence dates on the order of 9 mya are used, the mean rate for hominids becomes 0.5 to 1% per million years (Brown, 1985; Spuhler, 1988). Nei (1985) has also suggested a rate of approximately 0.7% per million years for mtDNA evolution, based on an estimated divergence time of 13 mya for Pongo. In addition, studies of mtDNA variability in several animal species have shown that gene flow between populations will obscure the effects of drift, making them appear to have diverged more recently than may actually be the case (Avise et al., 1986,1987; Ball et al., 1988; Graves et al., 1984). A slower rate of mtDNA evolution and the effects of gene flow between populations after divergence suggest that even if lineage extinction has not been a significant factor, the common origin of all modern human mtDNA can be traceable to the radiation of H. erectus out of Africa (see Wolpoff, 1989), rather than a recent radiation of modern populations.

In opposition to the views of Cann et al. (1987), Spuhler believes that the mtDNA evidence “. . . clearly supports the regional transition hypothesis . . .” (Spuhler, 1988: 42). We are certain that proponents of the “Out of Africa” view will not rally to Spuhler’s interpretation and that criticism of his appraisal will soon appear. Our point, however, is that there is still much disagreement among the experts about what the genetic data can and do tell us about modern human origins. Perhaps it will ultimately be conclusively shown that human genetic data do support an “Out of Africa” explanation of modern human origins, but at the present time this perspective is not unequivocally supported by these data as both the AES and RAE models assert.

CONCLUDING INTERPRETATIONS AND SPECULATIONS

We feel that the most important conclusion to be drawn from our present state of knowledge on modern human origins is that no model explains the available data on this subject unequivocally, whatever their proponents suggest. Each model has pros and cons, but overly dogmatic statements regarding when, where, and especially how modern humans made their debut are premature. Stating that we cannot be dogmatic, however, does not imply that we cannot confidently say anything about modern human origins. For example, it is clear that modern H. sapiens evolved from an archaic H. sapiens stock somewhere in the Old World. There is certainly no convincing evidence of any “Presapiens” lineages; and with the possible exception of the gracile group from Australasia, all of the earliest modern humans from each region exhibit archaic-reminiscent features, which demonstrate their descent from archaic H. sapiens forerunners. It is, of course, not clear to all that these features were derived, even in part, from different archaic forerunners in separate regions, in other words, that they represent clade features. This is certainly an area in need of detailed research. In a phenetic sense, however, these archaic-reminiscent features establish an archaic H. sapiens stage in the emergence of modern H. sapiens. The basic question yet to be answered is: Are these ancestral archaic H. sapiens restricted to only one region, or did the archaic population in the broad sense participate in modern human origins? We can summa- rize the issues pertinent to this basic question by posing four other questions.


When and where do modern humans first appear? Assuming that existing chronometric dates are correct, the earliest modern hu-

mans appear in Africa between 40 ky and 100 ky BP (130 ky BP if one accepts the highly questionable thoriuduranium mollusc shell date for Omo-Kibish Member l), in west Asia between 80 ky and 115 ky BP, in southern Europe around 50 ky BP (assuming the date on the Crete skeleton is correct), in central Europe between 34 ky and 36 ky BP, and in western Europe around 30 ky BP. Even if the very early ESR and TL dates for the earliest west Asians (QafzehiSkhul) are incorrect, an age of less than about 50 ky for these hominids seems highly unlikely. In Australasia, the date appears to be between 30 ky and 45 ky; and in east Asia proper, no modern hominid specimen is dated to >20 ky BP. In the last two regions, particularly east Asia, there are extensive temporal gaps between the latest archaic and earliest modern forms, and the fossil samples are so small that it is really uncertain when modern humans first appeared here.

The early dates for Africa and western Asia are all from relatively new chronometric techniques-at least new to the issue of modern human origins. We cannot help but wonder whether use of these techniques in Europe might push back the dates for initial modern hominids here as well. Thermoluminescence has, however, been used to date LeMoustier, and these results suggest that Neandertals were still in Europe between 46 ky and 60 ky BP (Valladas et al., 1986).

Is the origin of modern humans a classic monocentric phenomenon? Although it would appear that modern humans do appear late in Europe, it is not

clear if they emerged in one non-European location appreciably before they appeared in other regions. From the available paleoanthropological record, one can construct scenarios with an African, western Asian, and possibly even an East Asian homeland, depending on how the chronology is interpreted. The genetic evidence is equivocal here, since there are interpretations that would derive modern human mtDNA variability from African, Caucasoid, and Far Eastern stocks.

If the origin of modern humans was a classic monocentric, cladogenetic phenomenon, there should be evidence of a transition across the archaiclmodern human boundary in only one region: the moddrn homeland. We believe southern Africa exhibits evidence of such a transition, primarily owing to the nature of the Africa Transitional Group. However, there is little evidence of distinctly African clade features traversing the boundary. It is also our opinion that evidence of transition also occurs in other areas of the Old World. This evidence consists of fossil samples that are phenetically similar to the African Transitional Group and the presence of potential regionally specific clade features in certain regions.

In central Europe and East Asia for example, there are archaic hominids that exhibit many of the same general features, particularly in the face, that are cited as transitional features in the African Transitional Group. Perhaps, as some would argue, these are simply parallelisms; but it could also be argued that Africa and East Asia have patterns of facial reduction that occurred independently following the model proposed by Smith (1983, 1985). The same argument could conceivably also be made for central Europe too, but considering that this reduction is apparently later here than in Africa or East Asia, some extra-European influence is likely to be involved.

Our point here is not to argue that these phenomena actually occurred independently. We simply want to emphasize that reasonable transitional specimens and samples do exist outside Africa. If the definition of “transitional” were expanded to include archaic specimens or samples that approached the modern human condition to a greater extent than do earlier archaic humans in a particular region, then western Asia and Australasia could also be included in the list of regions with transitional samples. The catalyst for the development of these transitional mor- phologies may very well have involved gene flow from a single region, and Africa is perhaps the best regional candidate for this source. But if this is indeed the


explanation, southern Africa could obviously not have been isolated from the remainder of the world while the transition to modern humans (or perhaps more accurately, to modern human anatomical form) was taking place. Thus, this could not be a classic example of cladogenetic speciation even if the transition began in Africa.

Furthermore, there are indications of regionally distinct features of complexes that seem to extend across the archaiclmodern boundary in many non-African regions. Reasonable, although not overwhelming, cases can be made for such clade features in Europe, East Asia, and Australasia. Proponents of multiregional evolution have been routinely criticized for the use of phenetic patterns in arguing for regional continuity, but, it can be argued that there is more evidence for identifiable clade features in these regions than in western Asia and especially Africa. It is interesting to note that MRE would predict the appearance of clade features earliest a t the peripheries. Thus, while it may be true that supporters of multiregional evolution have overemphasized the evidence for regional clade features, we suggest that their opponents have too readily dismissed this evidence.

As one of us has previously stated (Smith, 1985), we believe that there is a certain logic to having a single region of origin for the genetic changes that ultimately resulted in modern human anatomical form. Acceptance of such a “center” is compatible with some versions of MRE as well as AES and RAE. On the basis of available paleoanthropological evidence, we are inclined to accept Africa as this center, even though we are skeptical of some of the evidence on which this interpretation is based. However, the evidence cited above suggests to us that a speci- ationheplacement model, such as RAE, is unlikely as an explanation for the emergence and radiation of modern human anatomical form. This is also indicated by the fact that any unequivocal definition of modern H. sapiens has to be, in part, regionally based. As Wolpoff (1986, 1990) has shown, a large percentage of obviously modern H. sapiens crania from Australia would have to be excluded from this taxon if some recent morphological definitions of the taxon (e.g., Day and Stringer, 1982) are used. Furthermore, many of the features that would exclude these modern Australians from modern H. sapiens are extensions of features that would differentiate the Australasian Ngandong sample from other archaic H. sapiens (Smith, 1987). If archaic Australasians had no input into modern Australasians, both of these observations become rather difficult to explain.

How did modern humans evolve? Some discussions of MRE have implied that the origin of modern humans in-

volved no novel genetic changes, only a recombination of genes already present in archaic H. sapiens. Obviously, in light of the great genetic similarity of chimpan- zees and humans, if there were genetic change associated with modern human origins, it could not have been extensive. We speculate that many of the changes that characterize late vis-a-vis early archaic H. sapiens groups in the western Old World may very well have involved a process of recombination of previously existing alleles, but that the final factor that led to modern human anatomical form may have been a change in the regulatory part of the genome. Specifically, we suggest a change in the control mechanism of endochondral bone growth, which resulted in shorter, more flexed anterior cranial bases and changes in the form of the long bones.

In Europe and western Asia, such a change would be commensurate with the differences that mark the archaiclmodern boundary and would explain the greater extent of change at the transition compared to periods preceding and succeeding it (Trinkaus and Smith, 1985). In Africa and the remainder of Asia, this is more difficult to evaluate because of the general absence of postcranial remains for archaic humans, and this of course makes our speculation even more problematic. However, it is not clearly contradicted by what we do know. In this context, the Jinniu Shan postcrania will be of great interest.

We believe, as Trinkaus (1983a,b, 1986, 1987) has argued, that the robusticity


characteristic of Neandertals (and presumably other archaic H. sapiens as well) was of considerable adaptive importance to them. Otherwise this extremely ener- getically expensive morphological pattern would not have been maintained so stringently in these hominids. Thus, before this complex could begin to reduce, the factors that necessitated the maintenance of this hyperrobusticity had to be less- ened or removed. We believe that the technological advances characteristic of the Middle Paleolithic (Mousterian, MSA, and Asian equivalents) provided the technological buffer that allowed such changes to begin. This is reflected in the early- to-late archaic H. sapiens trends of facial reduction discussed previously in several regional series. The regulatory shift that completed the reduction process was the final push across the archaiclmodern boundary. It is reasonable to think it probably occurred in only one area initially but rapidly spread because of its advantage in reducing energetic demands, both in terms of ontogeny and maintenance. The relative lateness of the appearance of this proposed regulatory change in Europe is problematic, assuming the African and western Asian dates are correct. There must have been something about the adaptive environment in Europe that re- quired the retention of hyperrobusticity much longer than in other regions, but exactly what this was is not clear.

Our speculative explanation takes a strongly “adaptivist” perspective on the issue of modern human origins, a perspective that is played down by the more replacement-oriented models. We feel, as Trinkaus and others (including the senior author) have urged, that the key to ultimately unraveling the puzzle of the origin of modern humans involves our ability to understand adaptively what the anatomical changes at the transition mean. We are a long way from accomplishing this difficult task, but such knowledge is crucial to understanding the emergence of modern human anatomical form (or H. sapiens, if speciation is involved).

Was the spread of modern human anatomical form a true migration? Although Brauer’s AES model, the recent synopsis by Cavalli-Sforza et al.

(19881, and perhaps even some proponents of the RAE model (Stringer, 1990) would allow for some influence of non-African archaic H. sapiens on modern humans in Eurasia, all emphasize that modern human populations radiated from Africa into Eurasia and replaced Eurasian archaic populations. While this may have been the case, we suspect that the emphasis on migration and replacement is too extreme. For one thing, there is really no compelling reason why African populations would have systematically expanded out of Africa a t this time. Our knowledge of population densities is minimal, but there is certainly no indication of population pressure or of marked changes in technology or subsistence efficiency until later (Klein, 1983). Environmental dessication is often cited as a potential catalyst for a radiation out of Africa, but no solid case for this has been offered. The one attempt to evoke environmental deterioration as the cause of modern humans spreading from North Africa into Europe (Boaz et al., 1982) has been systematically refuted (Fogarty and Smith, 1987). Thus, comparisons between the population dynamics revolving around the spread of food production and those related to the origin of modern humans (Cavalli-Sforza et al., 1988) are probably not appro- priate.

Wobst (1976) has demonstrated that populations with similar distributions and sizes to what must have been the case for Late Pleistocene humans would have had to maximize their contact with other groups to avoid extinction. This clearly would have demanded considerable movement of individual populations and would have resulted in extensive genetic exchange among contacting groups, perhaps even more than can be documented in extant hunter-gatherers. Such a situation could well explain the rather rapid spread of a strongly advantageous character (e.g., the reduction in hyperrobusticity) into regions where the local selective environment was suitable for its establishment. This process, although it would certainly include considerable localized population movement, would be better characterized


as assimilation rather than replacement or as demic diffusion with selection (Wol- poff, 1989) rather than systematic migration.

Obviously, we are of the opinion that a model that recognizes an important role for local continuity is the best explanation for modern human origins. Unlike some proponents of multiregional evolution, however, we are inclined to accept the idea that significant genetic change (as discussed above) was probably involved in the emergence of modern human anatomical form. Likewise, while we do not deny the theoretical possibility that this change could have occurred independently in different regions of the Old World (see previous discussion and Smith, 1985), present evidence suggests to us that it is more logical to view this change as having occurred initially in one region and then to have spread throughout the Old World. We differ with both “Out of Africa” models, however, in that we do not view this spread as ubiquitously resulting from population migration nor do we see local continuity as always playing the very minor role these models assert. We perceive our perspective on modern human origins to be theoretically closer to MRE than to the more replacement-oriented models. However, referring to it as MRE senso strict0 is somewhat misleading. We would describe our perspective as an assimilation model, since it involves assimilation of new elements into existing gene pools or, in some cases perhaps, old elements into new gene pools. It goes without saying that we believe such a model fits the paleoanthropological evidence better than other models, particularly those that emphasize replacement. Thus, we are not convinced that biological reality is best served by thinking of the origin of modern humans as the result of speciation in either a biological or evolutionary species concept sense. Since we do not view the origin of modern human anatomical form as a speciation event, we feel justified in continuing to use such terms as archaic H . sapiens and modern H . sapiens. While this explanation will certainly not satisfy those who find such usage unusual (Tattersall, 19861, we believe that they are adequate descriptive terms for dealing with temporal depth within a single, polytypic species.

Each of the models discussed in this review has contributed immeasurably to the enhancement of our understanding of later human evolution. However, we or anyone else involved in research on modern human origins must admit that there are still a number of fundamental questions to be answered and considerable gaps in our knowledge. Each of the models has its own particular strengths and weak- nesses, but none unequivocally explains all the available data. Thus, all of us should refrain from offering overly dogmatic or polemical interpretations on detailed aspects of modern human origins that are obviously not justifiable given our present state of knowledge regarding that phenomenon,

ACKNOWLEDGMENTS

We are grateful to our many colleages for discussions and information on various topics pertinent to this paper, particularly G. Brauer, M. Green, P. Habgood, G. Pope, J. Spuhler, E. Trinkaus, and M.H. Wolpoff. E. Szathmary deserves our thanks for her patience and assistance during the preparation of this manuscript. The authors would also like to thank M.O. Smith and S.E. Mitnick for their support and encouragement. Comments by three anonymous reviewers were also help- ful in preparing the final draft of this paper. The final manuscript was typed by L. Baradat. Many of the observations made in this paper are based on research supported by the National Academy of Sciences, National Science Foundation, Alex- ander von Humboldt Foundation, and the University of Tennessee.

LITERATURE CITED

Allsworth-Jones P (1986) The Szeletian and the Transition from Middle to Upper Palaeolithic in Central Europe. Oxford Clarendon Press.

Avise JC, Helfman GS, Saunders NC, and Hales LS (1986) Mitochondria1 DNA differentiation in North Atlantic eels: Population genetic come-

quences of an unusual life history pattern. Proc. Natl. Acad. Sci. USA 83:4350-4354.

Avise JC, Neigel JE , and Arnold J (1984) Demo- graphic influences on mitochondria1 DNA lineage survivorship in animal populations. J . Mol. Evol. 20:99-105.


Avise JC, Reeb CA, Saunders NC (1987) Geo- graphic structure and species differences in mitochondrial DNA of mouthbrooding marine cat- fishes (Ariidue) and demersal spawning of toadfishes (Butruchoidue). Evol. 41 :991-1002.

Bada J L [ 1987) Paleoanthropological applications of amino acid racemization dating of fossil bones and teeth. Anthropological Anzeiger 45:l-8.

Bada JL, and Deems L (1975) Accuracy of dates beyond the 14C dating limit using aspartic acid racemization reaction. Nature 255:218-219.

Ball RM, Freeman S, James FC, Bermingham E, and Avise JC (1988) Phylogeographic population structure of Red-winged Blackbirds assessed by mitochondrial DNA. Proc. Natl. Acad. Sci. USA 85:1558-1562.

Bartstra G-J, Soegondho S, and van der Wiik A (1988) Ngandong man: Age and artifacts. J. Hum. Evol. 17:325-337.

Bar-Yosef 0 (1990) Middle Palaeolithic chronology and the transition to the Upper Palaeolithic in Southwest Asia. In G Brauer and FH Smith (eds.): Continuity or Replacement: Controversies in Homo sapiens Evolution. Rotterdam: Balkema.

Beaumont PB, De Villiers H, Vogel J C (1978) Modern man in sub-Saharan Africa prior to 49000 years BP: A review and evaluation with particular reference to Border Cave. S. Afr. J . Sci. 74t409-419.

Beaumont PB (1980) On the age of the Border Cave Hominids 1-5. Palaeontol. Afr. 23:21-33.

Bellwood P (1985) Prehistory of the Indo-Malay- sian Archipelago. Sydney: Academic Press.

Binford L (1984) Faunal Remains from Klasies River Mouth. New York Academic Press.

Boas NT, Ninkovitch N, and Rossignol-Strick M (1982) Paleoclimatic setting for Homo supiens ne- underthulenszs. Natunvissenschaften 69:29-33.

Boule M (1921) Les Hommes Fossiles: filements de Paleontologie Humaine, 1st Ed. Paris: Masson et Cie.

Brauer G (1978) The morphological differentiation of anatomically modern man in Africa, with spe- cial regard to recent finds from East Africa. Z. Morphol. Anthropol. 69:266-292.

Brauer G (1980) Die morphologischen Aftinitaten des jungspleistozanen Stirnbeines aus dem Elb- mundungsgebiet bei Hahnofersand. Z. Morphol. Anthropol. 71:l-42.

Brauer G (1983a) Vom archaischen zum anato- misch modernen Homo sapiens in Afrika. Habi- litationsschrift. Fachbereich Biologie der Univer- sitat Hamburg.

Brauer G (1983b) Die menschlichen Skelettfunde des “Later Stone Age” aus der Mumba-Hohle und anderen Lokalitaten nahe des Evasi-Sees (Tanza- nia) und ihre Bedeutung fur die Populationsdif- ferenzierung in Ostafrika. In H Muller-Beck (ed.) Die Archaologischen und Anthropologischen Ergebnisse der Kohl-Larsen-Expeditionen in Nord-Tanzania 1933-1939. Tiibingen: Verlag Ar- chaeologica Venatoria Institut fur Urgeschicht der Universitat Tiibingen.

Brauer G (1984a) A craniological approach to the origin of anatomically modern Homo sapiens in Africa and implications for the appearance of modern Europeans. In FH Smith and F Spencer (eds.): The Origins of Modern Humans: A World Survey of the Fossil Evidence. New York: Alan R. Liss, Inc., pp. 327-410.

Brauer G (1984b) The Afro-european sapiens hy-

pothesis and hominid evolution is Asia during the Middle and Upper Pleistocene. In P Andrews and G Frazen (eds.): The Early Evolution of Man. Courier Forschungsinstitut Senckenberg, Vol. 69, pp. 145-165.

Brauer G (1984~) Presapiens-hypothese oder Afro- europaische Sapiens-hypothese? Z. Morphol. An- thropol. 75:l-25.

Brauer G (1989) The evolution of modern humans: A comparison of the African and non-African evidence. In P Mellars and CB Stringer (eds.): The Origins and Dispersal of Modern Humans: Be- havioural and Biological Perspectives. Edin- burgh: Edinburgh University Press.

Brauer G (1990) Africa’s place in the evolution of Homo sapiens. In G Brauer and FH Smith (eds.): Continuity or Replacement: Controversies in Homo sapiens Evolution. Rotterdam: Balkema.

Brauer G, and Leakey RE (1986) A new archaic Homo supiens skull from Eliye Springs, West Turkana, Kenya. Z. Morphol. Anthropol. 76:245- 252.

Brauer G, and Mehlman MJ (1988): Hominid mo- lars from a Middle Stone Age level at the Mamba Rock Shelter, Tanzania. Am. J . Phys. Anthropol.

Brauer G, and Smith FH, eds. (1990) Continuity or Replacement: Controversies in Homo supiens Ev- olution. Rotterdam: Balkema.

Brothwell DR (1960) Upper Pleistocene human skull from Niah caves Sarawak. Sarawak Mu- seum J . 9r323-349.

Brown WM (1980) Polymorphism in mitochondrial DNA of humans as revealed by restriction endonuclease analysis. Proc. Natl. Acad. Sci. 77:

Brown WM (1985) The mitochondrial genome of animals. In R MacIntyre (ed.): Molecular Evolu- tionary Genetics. New York: Plenum, pp 95-130.

Butzer K (1969) Geological interpretation of two Pleistocene hominid sites in the Lower Omo basin. Nature 222t1133-1135.

Butzer K (1976) The Mursi, Nkalabong and Kibish Formations, Lower Omo Basin, Ethiopia. In Y Coppens, FC Howell, GL Issac, and RE Leakey (eds.): Earliest Man and Environments in the Lake Rudolf Basin. Chicago: University of Chi- cago Press, pp. 12-23.

Butzer K, Beaumont PB, and Vogel JC (1978) Lithostratigraphy of Border Cave, Kwa Zulu, South Africa: A Middle Stone Age sequence be- ginning ca. 195,000 B.P. J . Archaeol. Sci. 5:317- 341.

Cann RL (1987) In search of Eve. The Sciences (SeptlOct):30-37.

Cann RL (1988) DNA and human origins. Annu. Rev. Anthropol. 17t127-143.

Cann RL. Stonekine M. and Wilson AC (1987) Mi-

7569 -76.

3605-3609.

tochondrial DNA ‘?nd human evolution. Nature 32.5~31-36 _ _ _ -~

Cavalli-Sforza L (1974) The genetics of human populations. Sci. Am. 231:81-89.

Cavalli-Sforza L, Piazza A, Menozzi P, and Moun- tain J (1988) Reconstruction of human evolution: Bringing together genetic, archaeological, and linguistic data. Proc. Natl. Acad. Sci. USA 85: 6002-6006.

Clark J D (1970) The Prehistory of Africa. New York: Praeger.

Clarke RJ (1985) A new reconstruction of the Flo- risbad cranium, with notes on the site. In E Del-


son (ed.): Ancestors: The Hard Evidence. New York: Alan R. Liss, Inc., pp 301-305.

Coon CS (1962) The Origin of Races. New York: Knopf.

Day MH (1969) Early Homo supiens remains from the Omo River Region of Southwest Ethiopia. Na- ture 222~1135-1138.

Day MH, and Stringer CB (1982) A Reconsidera- tion of the Omo Kibish Remains and the Erectus- Sapiens Transition. I. Cong. Int. PalBont. Hum. Nice, pp 814-846.

Day MH, Leakey MD, Magori C (1980) A new hominid fossil skull (L.H. 18) from the Nealoba Beds Laetoli, northern Tanzania. Nature 2g4:55- 56.

Deacon HJ (1989) Late Pleistocene palaeoecology and archaeology in the southern Cape. In P Mel- lars and CB Stringer (eds.): The Origins and Dis- persal of Modern Humans: Behavioural and Bio- logical Perspectives. Edinburgh: University of Edinburgh Press.

Deacon HJ, and Geleijnse VB (1988) The stratig- raphy and sedimentology of the main site sequence, Klasies River, South Africa. S. Afr. Arch. Bull. 43:5-14.

Deacon HJ, Geleijnse VB, Thackeray AI, Thack- eray JF, Tusenius ML, and Vogel JC (1986) Late Pleistocene cave deposits in the southern Cape: Current research at Klasies River. Palaeoecol. Africa 17t31-37.

Delson E (1988) One source not many. Nature 332: 206.

De Lumley MA, and Sonakia A (1985) Premiere decouverte d’un Homo erectus sur le continent In- dien a Hathnora, dans la moyenne vallee de la Narmada. L’Anthropologie 89:13-61.

De Villiers H (1973) Human skeletal remains from Border Cave, Ingwawuma District, KwaZulu, South Africa. Ann. Transvaal. Mus. 28:229-256.

De Villiers H (1976) A second adult mandible from Border Cave, Ingwawuma District, KwaZulu, South Africa. S. Afr. J. Sci. 72:212-215.

Diamond J (1989) The great leap forward. Dis- cover 10:50-60.

Eldredge N, and Tattersall I (1982) The Myths of Human Evolution. New York: Columbia Univer- sity Press.

Excoffier L, and Langaney A (1989) Origin and differentiation of human mitochondria1 DNA. Am. J . Hum. Genet. 44~73-85.

Excoffier L, Pellegrini B, Sanches-Mazas A, Simon C, and Langaney A (1987) Genetics and history of sub-Saharan Africa. Yrbk. Phys. Anthropol. 30: 151-194.

Facchini A, and Giusberti G (1990) Homo supiens supiens remains from the Island of Crete. In G Brauer and FH Smith (eds.): Continuity or Re- placement: Controversies in Homo supiens Evo- lution. Rotterdam: Balkema.

Fogarty ME, and Smith FH (1987) Late Pleis- tocene climatic reconstruction in North Africa and the emergence of modern humans. Hum. Evol. 2:311-319.

Fraver DW (1986) Cranial variation at Mladee and the relationship between Mousterian and Upper Paleolithic hominids. In VV Novotny and A Miz- erova (eds.): Fossil Man: New Facts, New Ideas. Papers in honor of J a n Jelinek’s Life Anniver- sary, Brno: Anthropos. 23~243-256.

Frayer DW (1990) The persistence of Neanderthal features in post-Neanderthal Europeans. In G

Brauer and FH Smith (eds.): Continuity or Re- placement: Controversies in Homo supiens Evo- lution. Roterdam: Balkema.

Giles E, and Ambrose SH (1986) Are we all out of Africa? Nature 322:21-22.

Glen E, and Kaczanowski K (1982) Human remains. In J K Koztowski (ed. Bacho Kiro Cave (Bulgaria saw: Panstwowe Wydawnictwo Naukowe, pp. 75-79.

Goede A, and Hitchman MA (1987) Electron spin resonance (ESR) analysis of marine gastropods from coastal archaeological sites in southern Af- rica. Archaeometry, 19:163-174.

Gould SJ (1987) Bushes all the way down. Nat. Hist. 87t12-19.

Gould SJ (1988) A novel notion of Neanderthal. Nat. Hist. 97:16-21.

Gould SJ (1989) Grimm’s greatest tale. Nat. Hist. 98t20-28.

Graves JE, Ferris SD, and Dizon AE (1984) Close genetic similarity of Atlantic and Pacific skipjack tuna (Katsuwonus pelamis) demonstrated with restriction endonuclease analysis of mitochondrial DNA. Mar. Biol. 79:315-319.

Green MK, and Groves CP (1989) A regional approach to the problem of the origin of modern humans in Australia. In P Mellers and CB Stringer (eds.): The Origins and Dispersal of Modern Humans: Behavioural and Biological Perspectives. Edinburgh: Edinburgh University Press.

Grine FE, and Klein RG (1985) Pleistocene and Holocene human remains from Equus Cave, South Africa. Anthropol. 8:55-98.

Habgood PJ (1988) A morphometric investigation into the origin(s) of anatomically modern humans. Dept. of Anthro. University of Sydney, Australia. Ph.D. Dissertation.

Habgood PJ (1989) The evolution of modern humans: Evidence from Australasia seen in a global context. In P Mellars and CB Stringer (eds.): The Origin and Dispersal of Modern Humans: Behav- ioral and Biological Perspectives. Edinburgh University of Edinburgh Press.

Habgood PJ (1990) The origins of anatomically modern humans in East Asia. In G Brauer and FH Smith (eds.): Continuity or Replacement: Controversies in Homo supiens Evolution. Rotter- dam: Balkema.

Harpending H (1974) Genetic structure of small populations. Annu. Rev. Anthropol. 3:229-243.

Hay RL (1987) Geology of the Laetoli area. In MD Leakey and JM Harris (eds.): Results of the La- etoli Expeditions 1975-1981. Oxford: Oxford University Press, pp 23-47.

Heim J (1982) Les Enfants NBandertalien de la Ferrassie. Paris: Foundation Singer-Polignac.

Hendey QB, and Volman TP (1986) Last intergla- cial sea levels and coastal caves in the Cape Prov- ince, South Africa. Quaternary Res. 25t189-198.

Hill AVS, and Wainscoat JS (1986): The evolution of the a- and b-globin gene clusters in human populations. Hum. Gen. 74:16-23.

Howell FC (1951) The place of Neanderthal man in human evolution. Am. J. Phys. Anthropol. 9: 379-416.

Howell FC (1957) The evolutionary significance of variation and varieties of “Neanderthal” man. Q. Rev. Biol. 32:330-410.

Howell FC (1978) Hominidae. In VJ Maglio and


HBS Cooke (eds.): Evolution of African Mam- mals. Cambridge: Harvard University Press, pp. 154-248.

Howells WW (1976) Explaining modern man: Ev- olutionists versus migrationists. J . Hum. Evol. 5,577-596.

Howells WW (1983) Origins of the Chinese people: Interpretations of recent evidence. In DM Keightly (ed.): The Origins of Chinese Civiliza- tion. Berkeley, pp. 297-319.

Hublin JJ (1988) Les plus anciens Representants de la LignBe, preneandertalienne. In: L'Homme de Neandertal, Vol. 3. L'Anatomie Liege: pp. 81- 94.

Hublin JJ, Tillier AM, and Tixier J (1987) L'humerus d'enfant Mousterien (Homo 4) du Jebel Irhoud (Maroc) dans son contexte arch& ologique. Bull. Mem. SOC. dAnthrop. 2:115-142.

Johnson MJ, Wallace DC, Ferris SD, Rattazzi MC, and Cavalli-Sforza LL (1983) Radiation of human mitochondria1 DNA types analyzed by restriction endonuclease cleavage patterns. J . Mol. Evol. 19: 255-271.

Kennedy GE (1984) The emergence of Homo supiens: The post cranial evidence. Man 19:94-110.

Kennedy GE (1990) The evolution ofHomo supiens as indicated by features of the post-cranium. In G Brauer and FH Smith (eds.): Continuity or Re- placement: Controversies in Homo supiens Evo- lution. Rotterdam: Balkema.

Kennedy KAR (1979) The deep skull of N iah An assessment of twenty years of speculation con- cerning its evolutionary significance. Asian Per- ~- spect. 20:32-50.

Kidder J H (1988) The transition from archaic Homo supiens to modern Homo supiens in Europe and the Near East proximal femoral evidence. Dept. of Anthropol. University of Tennessee. M.A. Thesis.

Klein RG (1983) The Stone Age prehistory of Southern Africa. Annu. Rev. Anthrouol. 12:25- 48.

Latorre A, Moya A, and Ayala FJ (1986) Evolution of Mitochondria1 DNA in Drosophila subobscuru. Proc. Natl. Acad. Sci. USA 8323649-8653.

Leakey MD, and Hay RL (1982) The chronological position of the fossil hominids of Tanzania. I. Cong. Int. Paleont. Nice, pp. 753-765.

Leroyer C, and Leroi-Gourhan A (1983) Problemes et chronologie: Le castelperronien et l'aurigna- cien. Bull. Soc. Prehist. Franc. 80:41-44.

Levbque F, and Vandermeersch B (1981) Le Nean- dertalien de Saint-Cesaire. Recherche 12: 242-244.

Levinton J (1988) Genetics, Paleontology, and Macroevolution. Cambridge: Cambridge Univer- sity Press.

Lewin R (1988) Modern humans under close scru- tiny. Science 239:1240-1241.

Lewin R (1989) Species questions in modern human origins. Science 243:1666-1667.

Lotka AJ (1931a) Population analysis-the extinction of families. I. J . Wash. Acad. Sci. 21:377- 380.

Lotka AJ (1931b) Population analysis-the extinction of families. 11. J . Wash. Acad. Sci. 21: 453-459.

Mayr E (1963) Animal Species and Evolution. Cambridge, MA: Belknap.

Mayr E (1969) Principles of Systematic Zoology. New York: McGraw Hill.

Mayr E (1982) The Growth of Biological Thought. Cambridge, MA: Belknap.

McCown TD, Keith A (1939) The Stone Age of Mount Carmel. 11: The fossil human remains from the Levalloiso-Mousterian. Oxford Claren- don Press.

Mehlman MJ (1987) Provenience, age, and associ- ations of archaic Homo supiens crania from Lake Eyasi, Tanzania. J . Arch. Sci. 14:133-162.

Mellars P, and Stringer CB, eds. (1989) The Origin and Dispersal of Modern Humans: Behavioural and Biological Perspectives. Edinburgh Edin- burgh University Press.

Minugh-Purvis N (1988) Patterns of craniofacial growth and development in Upper Pleistocene hominids. Dept. of Anthropol. University of Pennsylvania. Ph.D. Dissertation.

Morris AG (1990) Biological relationships between Upper Pleistocene and Holocene populations in Southern Africa. In G Brauer and FH Smith (eds.): Continuity or Replacement: Controversies in Homo Supiens Evolution. Rotterdam: Blakema.

Nei M (1985) Human evolution at the molecular level. In T Ohta and K Aoki (eds.): Population Genetics and Molecular Evolution. Tokyo: Japan Sci. SOC. Press, pp. 41-64.

Nei M and Roychoudhury AK (1982) Genetic Re- lationship and Evolution of Human Races. Evol. Biol. 14:l-49.

Oakley KP, Campbell BG, and Mollenson TI (1975): Catalogue of Fossil Hominids. Vol. 3, America, Asia, Australia. London: British Mu- seum (Natural History).

Parkington J (1989) A critique of the consen- sus view on the age of Howieson's Poort assemblages in South Africa. In P Mellars and CG Stringer (eds.): The Origins and Dispersal of Modern Humans: Behavioural and Biological Perspectives. Edinburgh: Edinburgh University Press.

Pope GG (1988) Recent Advances in Far Eastern paleoanthropology. Annu. Rev. Anthropol. 17: 43-77.

Pope GG, and Cronin J (1984) The Asian Homi- nidae. J . Hum. Evol. 13:377-396.

RadovEiC J (1988) Goljanovit-Kramberger and Krapina Early Man. Mladost: Zagreb.

Rak Y (1986) The Neanderthal: A new look at a n old face. J . Hum. Evol. 15:151-164.

Rak Y, and Arensburg B (1987) Kebara 2 Nean- derthal pelvis: First look at a complete inlet. Am. J . Phys. Anthropol. 23:227-231.

Rightmire GP (1978) Florisbad and human population succession in southern Africa. Am. J . Phys. Anthropol. 48:475-486.

Rightmire GP (1979) Implications of the Border Cave skeletal remains for later Pleistocene human evolution. Curr. Anthropol. 20.23-35.

Rightmire GP (1984) Homo supiens in sub- Saharan Africa. In FH Smith and F Spencer (eds.): The Origins of Modern Humans: A World Survey of the Fossil Evidence. New York: Alan R. Liss, Inc., pp. 295-325.

Rightmire GP (1985) The tempo of change in the evolution of mid-Pleistocene Homo. In E Delson (ed.): Ancestors: The Hard Evidence. New York Alan R. Liss, Inc., pp. 255-264.

Rightmire GP (1986) Africa and the origins of modern humans. In R Singer and J K Lundy (eds.): Variation, Culture and Evolution in Afri-


can Populations: Papers in Honor of Dr. Hertha de Villiers. Johannesburg: Witwatersrand Uni- versity Press, pp 209-220.

Rosenberg KR (1988) The functional significance of Neandertal pubic length. Curr. Anthropol. 29: 595 -61 7.

Santa Luca AP (1978) A re-examination of pre- sumed Neandertal fossils. J . Hum. Evol. 7.619- 636.

Santa Luca AP (1980) The Ngandong fossil hominids. Yale Univ. Pub. Anthropol. 78:l-175.

Schwarcz HP, Griin R, Vandermeersch B, Bar- Yosef 0, Vallades H, and Tchernov E (1988) ESR dates for the hominid burial site of Qafzeh in Is- rael. J. Hum. Evol. 17.733-737.

Simmons T, Falsetti AB, and Smith FH (in press) Evolutionary patterns in the frontal bones of Late Pleistocene hominids from Western Asia. J. Hum. Evol.

omy. New York: Columbia university Press. Simpson GG (1961) Principles of Animal Taxon-

Singer R, and Wymer JJ (1982) The Middle Stone Age at Klasies River Mouth in South Africa. Chi- cago: University of Chicago Press.

Smith FH (1982) Upper Pleistocene hominid evolution in South-Central Europe: A review of the evidence and analysis of trends. Curr. Anthropol. 23.667-703.

Smith FH (1983) Behavioral interpretations of changes in craniofacial morphology across the archaiclmodern Homo sapiens transition. In E Trinkaus (ed.): The Mousterian Legacy: Human Biocultural Change in the Upper Pleistocene. Br. Archaeol. Rep. Int. Ser. 164:141-163.

Smith FH (1984) Fossil Hominids from the Upper Pleistocene of Central Europe and the origins of modern Europeans. In FH Smith and F Spencer (eds.): The Origins of Modern Humans: A World Survey of the Fossil Evidence. New York Alan R. Liss, Inc., pp. 137-209.

Smith FH (1985) Continuitv and change in the origin of modern Homo sa&ns. Z . Morph. An- thropol. 75:197-222.

Smith FH (1987) Modern human origins: A perspective from the fossil record. Paper presented a t the annual meeting of the Geological Society of America. Phoenix, Arizona.

Smith FH (1990) The role of continuity in the origin of modern humans. In G Brauer and FH Smith (eds.): Continuity or Replacement Contro- versies in Homo Sapiens Evolution. Rotterdam: Balkema.

Smith FH and Paquette SP (1989) The adaptive basis of Neandertal facial form, with some thoughts on the nature of modern human origins. In E Trinkaus (ed.): The Emergence of Modern Humans: Biocultural Adaptations in the Later Pleistocene. Cambridge: Cambridge University Press.

Smith FH, and Raynard GC (1980) Evolution of the supraorbital region in Upper Pleistocene fossil hominids from South-Central Europe. Am. J . Phys. Anthropol. 53:589-610.

Smith FH, and Spencer F (eds.) (1984) The Origins of Modern Humans: A World Survey of the Fossil Evidence. New York: Alan R. Liss, Inc.

Smith FH, Simek J, and Harrill M (1989) Geo- graphical variation in supra-orbital torus reduction during the later Pleistocene (ca. 80,000- 15,000 BP). In P Mellars and CB Stringer (eds.): The Origins and Dispersal of Modern Humans:

Behavioural and Biological Perspectives. Edin- burgh: Edinburgh University Press.

Smith FH and Trinkaus E (in press) Modern human origins in central Europe: a case of continuity? In JJ Hublin and AM Tillier (eds.): Aux Orig- ines de la Diversite Humaine. Encyclopedia des Sciences et des Techniques. Paris: Presser Uni- versit6 de France.

Spencer F (1984) The Neandertals and their evolutionary significance: A brief historical survey. In FH Smith and F Spencer (eds.): The Origins of Modern Humans: A World Survey of the Fossil Evidence. New York: Alan R. Liss, Inc., pp. 1-49.

Spencer F (1986) Ecce Homo: An Annotated Bib- liographic History of Physical Anthropology. Connecticut: Greenwood Press.

Spencer F, and Smith FH (1981) The significance of Ale5 HrdliEka’s “Neanderthal phase of man”: A historical and current assessment. Am. J . Phys. Anthropol. 56:435-459.

Spuhler J N (1988) Evolution of mitochondrial DNA in monkeys, apes, and humans. Yrbk. Phys. Anthropol. 31~15-48.

Spuhler J N (in press) Evolution of mitochondrial DNA in human and other organisms. Am. J. Hum. Biol.

Stanlev SM (1979) Macroevolution. San Francisco: WH Freeman.

Stoneking M, and Cann RL (1989) African origin of human mitochondrial DNA. In P Mellars and CB Stringer (eds.): The Origins and Dispersal of Modern Humans: Behavioural and Biological Perspectives. Edinburgh University of Edin- burgh Press.

Stringer CB (1978) Some problems in Middle and Upper Pleistocene hominid relationships. In DJ Chivers and K Joysey (eds.): Recent Advances in Primatology, Vol. 3. Evolution. London: Aca- demic Press, pp. 395-418.

Stringer CB (1982) Towards a solution to the Ne- anderthal problem. J . Hum. Evol. I 1 ~431-438.

Stringer CB (1984) The definition of Homo erectus and the existence of the species in Africa and Eu- rope. Courier Forschungsinstitut Senckenberg 69:131-144.

Stringer CB (1988) The dates of Eden. Nature 331: 565-566.

Stringer CB (1989) Documenting the origin of modern humans. In E Trinkaus (ed.): The Emer- gence of Modern Humans: Biocultural Adapta- tions in the Later Pleistocene. Cambridge: Cam- bridge University Press.

Stringer CB (1990) Replacement, continuity, and the origin of Homo sapiens. In G Brauer and FH Smith (eds.): Continuity or Replacement: Contro- versies in Homo sapiens Evolution. Rotterdam: Balkema.

Stringer CB and Andrews P (1988) Genetic and fossil evidence for the origin of modern humans. Science 239.1263 -1268.

Stringer CB, and Trinkaus E (1981) The Shanidar Neandertal crania. In CB Stringer (ed.): Aspects of Human Evolution. London: Taylor and Fran- cis, pp. 129-165.

Stringer CB, Hublin J J , Vandermeersch B (1984) The origin of anatomically modern humans in Western Europe. In FH Smith and F Spencer (eds.): The Origins of Modern Humans: A World Survey of the Fossil Evidence. New York: Alan R. Liss, Inc., pp. 51-135.

Stringer CB, Griin R, Schwarcz HP, and Goldberg


P (1989) ESR dates for the hominid burial site of ES Skhul in Israel. Nature 338:756-758.

Tattersall I (1986) Species recognition in human paleontology. J . Hum. Evol. 15t165-175.

Tompkins R and Trinkaus E (1987) La Ferrassie 6 and the development of Neandertal pubic morphology. Am. J . Phys. Anthropol. 73:233-239.

Thorne AG (1976) Morphological contrasts in Pleistocene Australians. In RL Kirk and AG Thorne (eds.): The Origin of the Australians. Canberra: Australian Institute of Aboriginal Studies, pp. 95-112.

Thorne AG (1981) The centre and the edge: The significance of Australian hominids to African paleoanthropology. In RE Leakey and BA Ogot (eds.): Proceedings of the 8th Panafrican Con- gress of Prehistory and Quaternary Studies, Nairobi, September 1977. Nairobi: TILLMIAP, pp. 180-181.

Thorne AG, and Wolpoff MH (1981) Regional Con- tinuity in Australasian Pleistocene Hominid Ev- olution. Am. J . Phys. Anthropol. 55:337-349.

Trinkaus E (1982) A history of Homo erectus and Homo supiens paleontology in America. In F Spencer (ed.): “A History of American Physical Anthropology 1930-1980. New York: Academic Press, pp. 261-280.

Trinkaus E (1983a) The Shanidar Neandertals. New York: Academic Press.

Trinkaus E (1983b) Neandertal postcrania and the adaptive shift to modern humans. In E Trinkaus (ed.): The Mousterian Legacy: Human Biocul- tural Change in the Upper Pleistocene. Br. Ar- chaeol. Rep. Int. Ser. 164:165-200.

Trinkaus E (1984) Western Asia. In FH Smith and F Spencer (eds.): The Origins of Modern Humans: A World Survey of the Fossil Evidence. New York: Alan R. Liss, Inc., 251-293.

Trinkaus E (1986) The Neandertals and the origins of modern humans. Annu. Rev. Anthropol. 15:193-218.

Trinkaus E (1987) The Neandertal face: Evolu- tionary and functional perspectives on a recent hominid face. J. Hum. Evol. 16:429-443.

Trinkaus E, ed. (1989) The Emergence of Modern Humans: Biocultural Adaptations in the Later Pleistocene. Cambridge: Cambridge University Press.

Trinkaus E, and Smith FH (1985) The fate of the Neanderthals. In E Delson (ed.): Ancestors: The Hard Evidence. New York: Alan R. Liss, Inc., pp. 325-333.

Trinkaus E, and Thompson D (1987) Femoral dia- physeal histomorphometric age determinations for the Shanidar 3, 4, 5, and 6 Neandertals and Neandertal longevity. Am. J . Phys. Anthropol. 72:123-129.

Valladas H, Geneste JM, Joron JL, and Chadelle J P (1986) Thermoluminescence dating of Le Moust- ier (Dordogne, France). Nature 322:452-454.

Valladas A, Joron JL, Valladas G, Arensburg B, Bar-Yosef 0, Belfer-Cohen A, Goldberg P, Laville H, Meignen L, Rak Y, Tchernon E, Tillier AM, and Vandermeersch B (1987) Thermolumines- cence dates for the Neanderthal burial site at Ke- bara in Israel. Nature 330t159-160.

Valladas A, Reyes JL, Joron JL, Valladas, Bar- Yosef 0 (1988) Thermoluminescence dating of Mousterian ‘Proto-Cro-magnon’ remains from Is- rael and the origin of modern Man. Nature 331: 614-616.

Vallois HV (1958) La Grotte de Fontechevade. 2nd part: Anthropologie. Arch. Inst. Paleont. Hum. 29:l-164.

Vandermeersch B (1981a) Les Hommes fossiles de Qafzeh (Israel). Paris: CNRS.

Vandermeersch B (1981b) Les premiers Homo supiens au Proche-Orient. In D Ferembach (ed.) “Les Processus de YHominisation.” Paris: CNRS, pp. 97-100.

Vandermeersch B (1985) The origin of the Nean- derthals. In E Delson (ed.): Ancestors: The Hard Evidence. New York: Alan R. Liss, Inc., pp. 306- 309.

Vandermeersch B (1989) The evolution of modern humans: Recent evidence from South-west Asia. In P Mellars and CB Stringer (eds.): The Origins and Dispersal of Modern Humans: Behavioural and Biological Perspectives. Edinburgh: Edin- burgh University Press.

Van Valen LM (1986) Speciation and our own species. Nature 322:412.

Van Vark GN (1986) More on the classification of the Border Cave 1 skull. 5th Congress of the Eu- ropean Anthropological Associati&, Lisbon.

Vogel JC, and Beaumont PB (1972) Revised radiocarbon chronology for the Stone Age in South Af- rica. Nature 237:50-51.

Vrba E (1982) Biostratigraphy and chronology, based particularly on Bovidae, of southern hominid-associated assemblages: Makapansgat, Sterkfontein, Taung, Kromdraai, Swartkrans. I. Cong. Int. Paleontol. Hum. Nice, pp. 707-752.

Wainscoat JS, Hill AVS, Boyce AL, Flint J , Her- nandez M, Thein SL, Old JM, Lynch JR, Falusi AG, Wetherall DJ, and Clegg J B (1986) Evolu- tionary relationships of human populations from an analysis of nuclear DNA polymorphisms. Na- ture 319:491-493.

Weiss KM (1984) On the number of members of the genus Homo who have ever lived, and some evolutionary implications. Hum. Biol. 56:637- 649.

Weiss KM, and Maruyama T (1976) Archaeology, population genetics and studies of human racial ancestry. Am. J . Phys. Anthropol. 44:31-50.

Wendorf F, Laury RL, Albritton CC, Schild R, Haynes CV, Damon PE, Shafigullah M, Scarbor- ough R (1975) Dates for the Middle Stone Age of East Africa. Science 187:740-742.

Wiley EO (1981) Phylogenetics: The Theory and Practice of Phylogenetic Systematics. New York: John Wiley and Sons.

Wobst HM (1976) Locational relationships in paleolithic society. J. Hum. Evol. 5:49-58.

Wolpoff MH (1979) The Krapina dental remains. Am. J. Phys. Anthropol. 50:67-114.

Wolpoff MH (1980): Paleoanthropology. New York: Knopf.

Wolpoff MH (1985) Human evolution at the peripheries: The pattern at the eastern edge. In PV Tobias (ed.): Hominid Evolution: Past, Present, and Future. New York: Alan R. Liss, Inc., pp 355-365.

Wolpoff MH (1986) Describing anatomically modern Homo supiens: A distinction without a defin- able difference. In VV Novotny and A Mizerova (eds.): Fossil Man: New Facts, New Ideas. Papers in honor of J a n Jelinek’s Life Anniversary. Brno: Anthropos, 23:41-53.

Wolpoff MH (1989) Multiregional evolution: The fossil alternative to Eden. In P Mellars and CB


Stringer (eds.): The Origins of Modern Humans: Behavioural and Biological Perspectives. Edin- burgh: Edinburgh University Press.

Wolpoff MH (1990) Theories of modern human origins. In G Brauer and FH Smith (eds.): Continu- ity or Replacement Controversies in Homo supiens Evolution. Rotterdam: Balkema.

Wolpoff MH, Smith FH, Malez M, RadovEiC J , Rukavina D (1981) Upper Pleistocene hominid remains from Vindija Cave, Croatia, Yugoslavia. Am. J . Phys. Anthropol. 54:499-545.

Wolpoff MH, Spuhler JN, Smith FH, RadovEiC J, Pope G, Frayer D, Eckhardt R, and Clark G (1988) Modern human origins. Science 241:772-773.

Wolpoff MH, Wu Xinzhi, and Thorne AG (1984) Modern Homo supiens origins: A general theory of hominid evolution involving the fossil evidence from east Asia. In FH Smith and F Spencer (eds.): The Origins of Modern Humans: A World Survey of the Fossil Evidence. New York: Alan R. Liss, Inc., pp. 411-483.

Wu R, and Dong X (1985) Homo erectus in China.

In Wu Rukang and JW Olsen (eds.): Palaeoan- thropology and Palaeolithic Archaeology in the People’s Republic of China. Orlando: Academic Press, pp. 78-89.

Wu X, and Wu M (1985) Early Homo supiens in China. In Wu Rukang and JW Olsen (eds.): Palaeoanthropology ana Palaeolithic Archaeol- ogy in the People’s Republic of China. Orlando: Academic Press, pp. 91-96.

Wu X, and Zhang Z (1985) Homo sapiens remains from late Palaeolithic and Neolithic China. In Wu Rukang and JW Olsen (eds.): Palaeoanthro- ~o loev and Palaeolithic Archaeoloev in the Peo- ple’TRepublic of China. Or lanz : Academic Press, pp. 107-133.

Yasuda N, Cavalli-Sforza LL, Skolnick M, and Mo- roni A (1974) The evolution of surnames: An analysis of their distribution and extinction. The- oret. Popul. Biol. 5~123-142.

Zhou M, Li Y, Wang L (1982) Chronology of the Chinese fossil hominids. I. Cong. Int. Paleont. Hum. Nice, pp. 593-604.

Modern human origins

Documents