1. Sally McBrearty The revolution that wasn’t: a new
Department of Anthropology, interpretation of the origin of modern
University of Connecticut, human behavior
Storrs, Connecticut 06269,
U.S.A. E-mail:
Proponents of the model known as the ‘‘human revolution’’ claim
mcbrearty@uconn.edu
that modern human behaviors arose suddenly, and nearly simul-
taneously, throughout the Old World ca. 40–50 ka. This fundamental
Alison S. Brooks behavioral shift is purported to signal a cognitive advance, a possible
Department of Anthropology, reorganization of the brain, and the origin of language. Because the
George Washington earliest modern human fossils, Homo sapiens sensu stricto, are found in
University, Washington, Africa and the adjacent region of the Levant at >100 ka, the ‘‘human
DC 20052, U.S.A. E-mail: revolution’’ model creates a time lag between the appearance of
abrooks@gwu.edu anatomical modernity and perceived behavioral modernity, and
creates the impression that the earliest modern Africans were behav-
Received 3 June 1999 iorally primitive. This view of events stems from a profound Euro-
Revision received 16 June centric bias and a failure to appreciate the depth and breadth of the
2000 and accepted 26 July African archaeological record. In fact, many of the components of
2000 the ‘‘human revolution’’ claimed to appear at 40–50 ka are found in
the African Middle Stone Age tens of thousands of years earlier.
Keywords: Origin of Homo These features include blade and microlithic technology, bone tools,
sapiens, modern behavior, increased geographic range, specialized hunting, the use of aquatic
Middle Stone Age, African resources, long distance trade, systematic processing and use of
archaeology, Middle pigment, and art and decoration. These items do not occur suddenly
Pleistocene. together as predicted by the ‘‘human revolution’’ model, but at sites
that are widely separated in space and time. This suggests a gradual
assembling of the package of modern human behaviors in Africa, and
its later export to other regions of the Old World. The African Middle
and early Late Pleistocene hominid fossil record is fairly continuous
and in it can be recognized a number of probably distinct species that
provide plausible ancestors for H. sapiens. The appearance of Middle
Stone Age technology and the first signs of modern behavior coincide
with the appearance of fossils that have been attributed to H. helmei,
suggesting the behavior of H. helmei is distinct from that of earlier
hominid species and quite similar to that of modern people. If on
anatomical and behavioral grounds H. helmei is sunk into H. sapiens,
the origin of our species is linked with the appearance of Middle
Stone Age technology at 250–300 ka.
2000 Academic Press
Journal of Human Evolution (2000) 39, 453–563
doi:10.1006/jhev.2000.0435
Available online at http://www.idealibrary.com on
Introduction and background 2000; Diamond, 1992; Mellars, 1995,
1996; Nobel Davidson, 1991; Tattersall,
The human revolution in Europe 1995; Bar-Yosef, 1998). The ‘‘human
For at least the past 15 years, most recon- revolution’’ model proposes a dramatic
structions of later human evolutionary alteration in human behavior at the Middle
history have featured a relatively brief and Paleolithic to Upper Paleolithic transition at
dramatic shift known as the ‘‘human revol- about 40 ka. This behavioral breakthrough
ution’’ (Binford, 1985, 1989; Mellars is thought by some to correspond to
Stringer, 1989; Klein, 1989a, 1994, 1995, increased cognitive sophistication, the
0047–2484/00/110453+111$35.00/0 2000 Academic Press

2. 454 .   . . 
manipulation of symbols, and the origin of two great European wars, and the trend
language (e.g., White, 1982; Mellars in archaeology in the second half of the
Stringer, 1989; Diamond, 1992; Byers, twentieth century has been the study of
1994; Mithen, 1994, 1996; Klein, 1995; but local sequences and the application of
see Kay et al., 1998). We believe that the models of cultural evolution (Otte Keeley,
model of the ‘‘human revolution’’ is fatally 1990).
flawed. Modern humans and modern In terms of developments in world pre-
human behaviors arose first in Africa, and history, however, Western Europe is a
we examine the African record to reveal a remote cul de sac with a somewhat anom-
different picture of the nature of events. alous prehistoric record. We argue here that
The concept of a ‘‘human revolution’’ and models derived from the unique record of
the periodization of Stone Age prehistory European prehistory do not explain events
have their roots in the nineteeth-century in Africa where the origin of modern people
probings of the Western European archaeo- actually occurred. In the Holocene, western
logical record. The first paleolithic classifi- Europe experienced a series of incursions
catory schemes were based on the Western from the less peripheral portions of the Old
European large mammal succession (Lartet World. Each arrival of a wave of invaders
Christy, 1865–1875; Lyell, 1868), and and alien technology induced a fairly
these authors emphasized the wide techno- sudden, rapid cultural turnover. These
logical gulf separating the l’age du renne disruptive episodes are reflected in the
(Upper Paleolithic) from the earlier phases European archaeological record as dis-
(Lartet Christy, 1865–1875:25). By the continuities that punctuate industrial
1920s the concept of an Upper Paleolithic periods of relatively long duration. They
distinguished by the appearance of engrav- have been sometimes described as ‘‘revol-
ing, sculpture, painting, beads, and worked utions,’’ such as the ‘‘neolithic revolution’’
bone tools had become current. A tripartite of Childe (1936, 1942).
division into Lower, Middle and Upper Recent paleoclimatic data and refined
Paleolithic based upon stone tool technology chronologies have supported the early sug-
(De Mortillet, 1900; Obermeier, 1924; gestion of Howell (1951) that regions of
Burkitt, 1921, 1928, 1933; Kendrick, 1925; Pleistocene Europe were repeatedly isolated
Menghin, 1931) echoed the three-age sys- by ice and mountain barriers, so that its
tems of Thomsen (1837) and Worsaae hominid populations were periodically
(1849) that partitioned the total prehistoric reduced or even eliminated (Howell, 1952;
record into ages of Stone, Bronze, and Iron. Gamble, 1986, 1994; Jochim, 1987;
The Lower, Middle and Upper Paleolithic Hublin, 1998a). Moreover, it has become
divisions of the western European record increasingly clear that the Neanderthals
have continued to dominate discourse in the were replaced by modern humans in Europe
field, despite problems in the application of within too short a period for the former to
these divisions to sequences in Eastern and have evolved into the latter (Mellars,
Southern Europe (Morselli, 1926:292). 1998a,b, 1999, Bocquet-Appel Demars,
None of these temporal divisions was 2000). Thus, the ‘‘revolutionary’’ nature of
intended as an evolutionary scheme, but the European Upper Paleolithic is most
rather they were thought to reflect repeated probably due to discontinuity in the
invasions by outsiders with new ideas. archaeological record rather than to the sort
Perhaps not surprisingly, a picture of of rapid cultural, cognitive, and/or biological
Europe conquered by invaders with superior transformation that has been argued by
technology had little appeal in the light of proponents of the ‘‘human revolution.’’

3.    ’ 455
The earliest modern Europeans were The fossil evidence for an African origin
Africans for modern humans is robust. It is clear that
modern humans (H. sapiens sensu stricto)
Who were the earliest modern Europeans? It were certainly present in Africa by 130 ka
is becoming increasingly difficult to deny (Day Stringer, 1982; Deacon, 1989), and
that they were Africans. Although the perhaps as early as 190 ka if specimens
‘‘mitochondrial Eve’’ hypothesis, first such as Singa are considered modern
articulated by Cann et al. (1987), has been (McDermott et al., 1996; Stringer, 1996).
revised in light of criticism (Templeton, Modern humans do not appear in Europe or
1992; Hedges et al., 1992; Ayala, 1995), and Central Asia before ca. 40 ka; earliest dates
population size and structure have effects on for the Levant range between ca. 80 ka and
the distribution of genetic characters that 120 ka (Day, 1969, 1972; Day Springer,
were not taken into account in early recon- 1982, 1991; Stringer, 1989, 1992;
structions (Harpending et al., 1993, 1998; McBrearty, 1990b; Stringer et al., 1989;
Sherry et al., 1994; Relethford, 1995; Brauer, 1984a,b, 1989; Stringer Andrews,
¨
Relethford Harpending, 1995), genetic 1988; Valladas et al., 1988; Grun ¨
data either directly support or are consistent Stringer, 1991; Miller et al., 1991; Foley
with an African origin for modern humans Lahr, 1992; Mercier et al., 1993; Deacon,
(Wainscoat et al., 1986; Cann, 1988; 1993b; Brooks et al., 1993a,b; Stringer,
Stringer Andrews, 1988; Vigilant et al., 1993a; Schwarcz, 1994; Straus, 1994;
1991; Stoneking, 1993; Stoneking et al., Bar-Yosef, 1994, 1995a, 1998; but see
1993; Relethford Harpending, 1994; Howells, 1989). Recent evidence suggests
Ayala, 1995; Nei, 1995; Goldstein, 1995; that modern humans were present in
Tishkoff et al., 1996; Ruvolo, 1996, 1997; Australia as early as 62 ka (Stringer, 1999;
Irish, 1998; Pfeiffer, 1998; Zietkiweicz et al., Thorne et al., 1999).
1997; Pritchard et al., 1999; Quintana- Although some, notably Brauer (1984a,b,
¨
Murci, 1999; Relethford Jorde, 1999; 1989), favor a scenario involving some inter-
Tishkoff et al., 2000; see Relethford, 1998 breeding among Neanderthal and modern
and Jorde et al., 1998 for recent reviews). human populations, the successful extrac-
As Howell (1994:306) observes, ‘‘The tion and analysis of fragmentary mito-
phylogenetic roots of modern humans are chondrial DNA (mtDNA) from both the
demonstrably in the Middle Pleistocene. Neanderthal type fossil (Krings et al., 1997,
The distribution of those antecedent 1999) and additional material from the
populations appear to lie outside of western northern Caucasus (Ovchinnikov et al.,
and eastern Eurasia, and more probably 2000) appears to remove the Neanderthals
centered broadly on Africa.’’1 from modern human ancestry. Body propor-
tions of early European H. sapiens fossils
1. The Middle to Late Pleistocene boundary is the suggest a tropical adaptation and support an
beginning of the last interglacial, at approximately African origin (Holliday Trinkaus, 1991;
130 ka; the base of the Middle Pleistocene is the shift
from reversed to normal magnetic polarity at the Ruff, 1994; Pearson, 1997, 2000; Holliday,
Matuyama–Brunhes boundary, dated to about 780 ka 1997, 1998, 2000). A single migration or
(Butzer Isaac, 1975; Imbrie Imbrie, 1980; Berger population bottleneck was originally envis-
et al., 1984; Martinson et al., 1987; Shackleton et al.,
1990; Deino Potts, 1990; Cande Kent, 1992; aged in the ‘‘African Eve hypothesis’’ (Cann
Baksi et al., 1992; Tauxe et al., 1992). Further evidence et al., 1987), but a succession of population
may confirm recent suggestions (Schneider et al., 1992; dispersals, subsequent isolation induced by
Singer Pringle, 1996; Hou et al., 2000) that the age
of this geomagnetic polarity reversal be revised to climatic events and local adaptation may
ca. 790 ka. better account for the complexity of the

4. 456 .   . . 
fossil record and the genetic composition of rather than by genetic processes, the most
present human populations (Howells, 1976, likely scenario would be an accretionary
1989, 1993; Boaz et al., 1982; Foley Lahr, process, a gradual accumulation of modern
1992; Lahr Foley, 1994, 1998; Ambrose, behaviors in the African archaeological
1998b). record (cf. Allsworth-Jones, 1993). This
It can be deduced from the archaeological change need not be unidirectional or con-
evidence that on a continent-wide scale the fined to a single location. Rather, we might
African record differs markedly from that expect innovative behaviors to appear at
of Europe in its degree of population con- different times and in different regions, and
tinuity. While parts of Africa, such as the due to low population densities we might
Sahara or the interior of the Cape Province expect the transmission of new ideas to be
of South Africa, do appear to have experi- sporadic.
enced interruptions in human settle- As early as the 1920s it was clear that the
ment during glacial maxima (Deacon African archaeological record could not be
Thackeray, 1984; Williams, 1984; Butzer, accommodated within the European Paleo-
1988b; Brooks Robertshaw, 1990; lithic model. A separate scheme of Earlier,
Mitchell, 1990), climatic reconstructions Middle and Later Stone Ages (ESA, MSA,
suggest that the contiguous expanse of and LSA) was devised for Stone Age Africa
steppe, savanna and woodland biomes avail- (Goodwin van Riet Lowe, 1929) to
able for human occupation, especially in the emphasize its distinctiveness from the
tropical regions of the continent, was always Lower, Middle, and Upper Paleolithic of
substantially larger than the comparable Europe. The ESA, MSA and LSA were first
regions in Europe. Perhaps as a result, defined on technological grounds on the
hominid populations in Africa, while prob- basis of material from South Africa
ably widely dispersed, appear to have (Goodwin, 1928; Goodwin van Riet
been consistently larger (Relethford Lowe, 1929). The terms were formally
Harpending, 1995; Jorde et al., 1998; endorsed by the Panafrican Congress of
Relethford Jorde, 1999; Tishkoff et al., 1955 (Clark, 1957a: xxxiii). The ESA as it is
2000). now understood includes both the Oldowan
and the Acheulian; the MSA encompasses
flake and blade tool industries which often
Revolution or evolution? The African
include prepared cores and points; and the
data
LSA is characterized by microlithic tech-
How might the archaeological signature of nology. The MSA was distinguished by the
continuous evolutionary change be expected presence of prepared core technology and,
to differ from that of abrupt replacement? If at most sites, unifacial and/or bifacial projec-
the entire human species experienced a sim- tile points, and by the absence of handaxes
ultaneous, punctuated, genetically encoded and microliths, hallmarks of the Acheulian
event, such as the development of modern and LSA respectively.
capacities for language (Klein, 1995; Before 1972, in the absence of accurate
Diamond, 1992), one would expect the chronometric dates, a radiocarbon date of
transition to modern human behavior to be 60 ka from Acheulian levels at Kalambo
abrupt, in Africa as well as in Europe and Falls, Zambia (Clark, 1969) was not
Asia. On the other hand, if aspects of recognized as infinite. This frequently cited
modern human culture in Africa were devel- date was particularly influential in establish-
oped by hominids using existing cognitive ing the impression of a short chronology
capabilities and transmitted by cultural for Africa. The MSA, at 60 ka, was

5.    ’ 457
considered the temporal equivalent of the quently retired (Bishop Clark, 1967:
Upper Paleolithic of Europe. Therefore the 987), when a mixture of different occu-
discovery of anatomically modern human pation levels was found to have occurred
remains associated with MSA artefacts at during excavation at the ‘‘Second Inter-
the South African sites of Border Cave mediate’’ type site of Magosi (Wayland
and Klasies River occasioned no surprise.2 Burkitt, 1932; Clark, 1957b; Hole, 1959;
The degree of regional differentiation, the Cole, 1967). Yet anachronisms, as well as
ubiquitous presence of blades and blade long periods of transition between stages,
cores, and the sophistication of projectile remain as problems (Vishnyatsky, 1994).
point technology in the African MSA were A fairly abrupt MSA–LSA transition is
considered comparable to the European apparent in the Mediterranean zones at the
Upper Paleolithic. However, bone tools, art northern and southern margins of Africa.
objects and beads were sparse when com- This seems consistent with the significant
pared to the European Upper Paleolithic, documented gaps in the settlement history
particularly the late Upper Paleolithic. of both regions (Close et al., 1990; Wendorf
The rarity of elements regarded as critical et al., 1990, 1993a; Mitchell, 1990; Deacon
to modern human culture in the MSA Thackeray, 1984; Klein, 1989b: 307).
served as grounds for regarding Africa as a However, at rock shelter sites in tropical
‘‘cultural backwater,’’ the place that initially Africa with relatively continuous occu-
gave rise to humanity, but failed to nurture pational records, such as Mumba, Tanzania
its later development (e.g., Butzer, 1971; (Mehlman, 1979, 1989), Matupi, D. R.
cf. Clark, 1975). In the later 1970s, new Congo (van Noten, 1977) and White
dating techniques and more accurate Paintings, Botswana (Robbins Murphy,
climatic correlations pushed back the age of 1998; Robbins et al., under review) there is a
the MSA well beyond 100 ka. The MSA was gradual transition from MSA to LSA tech-
recognized as the temporal equivalent of the nology over as much as 30 ka. Mehlman
European Middle Paleolithic, not the Upper (1991) has urged the development of new
Paleolithic. Attention focused on the human paradigms to accommodate the lack of a
fossils associated with the MSA, which were punctuated event.
now thought to be anomalously modern Because of the late, sudden, and nearly
in appearance. The fact that many MSA simultaneous appearance in Europe of mod-
artefacts recalled the Upper Paleolithic of ern humans and complex behavior, archae-
Europe in both form and technology was ologists working in Africa have sought a
forgotten. similar ‘‘human revolution’’ there. The fully
The use of a classifactory scheme developed signature of modern human
designed for Africa did not entirely remove behavior, including planning, sophisticated
ambiguity, as many industries displayed technology and resource use, and symbolic
characteristics of two different stages and behavior in the form of decorative art is
could not be assigned to one of the three clearly present in the African LSA. As a re-
divisions. Long transitional periods or sult, the MSA–LSA transition has been con-
‘‘Intermediates’’ were added to the tripartite flated with the Middle to Upper Paleolithic
ESA–MSA–LSA scheme at the 1955 and the emergence of modern human behav-
Panafrican Congress (Clark, 1957a: xxxiii), ior. Consequently the earliest anatomically
but the ‘‘Intermediate’’ concept was subse- modern humans, which occur in MSA con-
texts, are not accepted as fully ‘‘human’’.
2. The name Klasies River rather than Klasies River
Mouth or KRM is adopted here to conform with the We suggest that the expectation of a
recent usage of Hilary Deacon and his team. ‘‘human revolution’’ in Africa is ultimately

6. 458 .   . . 
a misapplication of a European model. the genus Homo when compared to the more
Further, we reject the idea of a time lag derived state in the Neanderthals (Rak,
between anatomical and behavioral change 1993).
in Africa, such as that proposed by Klein Specimens formerly attributed to
(1992, 1994, 1995, 1998). There was no ‘‘archaic’’ H. sapiens exhibit a number of
‘‘human revolution’’ in Africa. Rather, in plesiomorphic traits, including long low
this paper we present data from the human crania, large brow ridges, large, prognathic
fossil and archaeological records to show faces with large teeth, and the lack of a chin.
that novel features accrued stepwise. The chief justification for the inclusion of
Distinct elements of the social, economic, these fossils in our species has been their
and subsistence bases changed at different large brain sizes, though brain size is in part
rates and appeared at different times and a function of body mass, known to be quite
places. We describe evidence from the large among these hominids (Grine et al.,
African MSA to support the contention that 1995; Ruff et al., 1997; Kappelman, 1997).
both human anatomy and human behavior Recent discussions of later hominid phy-
were intermittently transformed from an logeny (e.g., Stringer, 1993b, 1994, 1995,
archaic to a more modern pattern over a 1996; Lahr Foley, 1994; Foley Lahr,
period of more than 200,000 years. 1997; Rightmire, 1998) have recognized the
distinctiveness of non-Neanderthal Middle
Pleistocene hominids and have resurrected
The hominid fossil record
the taxon H. heidelbergensis Schoetensack,
Until recently, most reconstructions of later 1908 for them, but we question the
human phylogeny recognized only one attribution of the African material to this
species after H. erectus. Grade-based taxon.
schemes commonly divided H. sapiens into Paradoxically, H. sapiens Linnaeus, 1758
two variants, ‘‘archaic’’ H. sapiens and ‘‘ana- lacks a satisfactory definition. Howell (1978:
tomically modern’’ H. sapiens (H. sapiens 201) observed over 20 years ago,
sensu stricto). The Neanderthals were then
‘‘The extensive relevant literature reveals an
sometimes distinguished from other unexpected lack of concern with the bio-
‘‘archaic’’ H. sapiens at the subspecific logical distinctiveness of a now-dominant
level as H. sapiens neanderthalensis (e.g., mammalian species’’,
Campbell, 1964). We concur with such and the situation is virtually unchanged
authors as Tattersall (1986, 1992), Kimbel today. The anatomy of H. sapiens is charac-
(1991), Harrison (1993), Rak (1993) and terized by a high round cranium, a chin, a
Stringer (1994, 1996) that there are grounds small orthognathic face, as well as reduced
for distinguishing ‘‘archaic’’ from ‘‘modern’’ masticatory apparatus and brow ridges. It
H. sapiens at the species level, and thus has been argued that most of these features
we regard the appearance of ‘‘modern’’ H. can be explained by greater flexion in the
sapiens as a speciation event. Here, we treat basicranium of H. sapiens (Lieberman,
the Neanderthals as the distinct species H. 1998b; Spoor et al., 1999).
neandertalensis King, 1864, and use the Because early fossils of H. sapiens dating
name H. sapiens to refer only to H. sapiens to 130 ka, and perhaps as early as 190 ka,
senus stricto. The use of the ‘‘anatomically are found in Africa (Grun et al., 1990;
¨
modern’’ label for H. sapiens sensu stricto is Deacon, 1989, 1993b; Day Stringer,
not only unnecessary but also misleading, as 1982; McDermott et al., 1996), it is reason-
many of the cranial features used to dis- able to seek evidence for the processes lead-
tinguish H. sapiens are in fact primitive for ing to the origin of H. sapiens in the African

7.    ’ 459
record of the Middle Pleistocene (Howell, works (Brauer, 1984a,b, 1989; Smith, 1985,
¨
1994). Although often described as 1993; Clark, 1988; Klein, 1989b, 1994;
‘‘scrappy’’ or insubstantial, the African Stringer, 1993a). Data in Table 1 roughly
hominid fossil sample from this time period follow Day’s tripartite construct, though this
numbers several dozen individuals (Table 1, should not be construed as an endorsement
Figure 1). While the circumstances of for anagenesis or a grade-based taxonomy.
recovery for some of the specimens are far Both the ascription of fossils to group and
from ideal, this is unfortunately true for the attachment of taxonomic labels are
many fossil discoveries, and in fact a fair problematic, and Group 1 specimens
number of the African specimens were probably belong to several different species
recovered by controlled excavation (e.g., (e.g., H. louisleakeyi, H. rhodesiensis).
Ndutu, Cave of Hearths, Haua Fteah, The principal unresolved issue in the
Mumba, Ngaloba, Klasies, Kapthurin post- clarification of the evolutionary relationships
cranials). For others, stratigraphic context of the hominids in Group 1 is the enigmatic
can be reasonably inferred, despite the fact status of H. erectus. This species is believed
that they are surface finds (e.g., Kapthurin by many to have been confined to Asia
mandibles, Eyasi, Buia). (Andrews, 1984; Tattersall, 1986; Groves,
Howell (1994: 305f) has deemed the sol- 1989; Clarke, 1990; Kimbel, 1991; Larick
ution of the evolutionary relationships Ciochon, 1996). Following Wood (1991,
among later Middle Pleistocene hominid 1992), some now ascribe to H. ergaster
populations one of the central problems in African fossils in the 1·5–2 Ma age range
the study of human evolution, and the taxo- formerly attributed to H. erectus. Other
nomic status of the African fossils is much authors (e.g., Rightmire, 1990, 1994, 1995,
debated (e.g., Tattersall, 1986; Clarke, 1998; Brauer Mbua, 1992; Harrison,
¨
1990; Foley, 1991a; Kimbel, 1991; 1993; Walker, 1993; Brauer, 1994) regard
¨
Stringer, 1992, 1993a, 1994, 1996; Aiello, H. erectus as a single polytypic species dis-
1993; Foley Lahr, 1997; Lahr Foley, tributed throughout most of the Old World,
1998; Rightmire, 1998). Revision of fossils and African specimens in our Group 1,
ascribed to Homo (Wood, 1991, 1992; spanning a broad range of time, continue
Wood Collard, 1999) has resulted in a to be ascribed to this taxon (e.g., OH9,
more ‘‘bushy’’ or speciose taxonomic Kapthurin, in Wood, 1992). Attribution of
picture for our genus in the Pliocene and the African and Asian Middle Pleistocene
Early Pleistocene, but for the African material to a single species assumes an
Middle and Late Pleistocene a unilineal adequate degree of gene flow to prevent
model is often invoked. It is our belief that speciation, but the archaeological differ-
the number of African Middle and Late ences between the regions suggest long term
Pleistocene hominid species has been under- isolation (Schick, 1994; but see Hou et al.,
estimated, because behavioral and repro- 2000).
ductive isolation may precede changes in the Group 1 in our scheme includes the
bony skeleton (Tattersall, 1986, 1992, Kabwe (Broken Hill) cranium, type speci-
1993; Rak, 1993). men of H. rhodesiensis Woodward, 1921, as
Over 25 years ago, Day (1973) suggested well as OH9. The latter specimen is usually
separating African Middle and Late referred in the literature to H. erectus, but
Pleistocene hominids into ‘‘early,’’ ‘‘inter- Louis Leakey (1961, 1963) emphatically
mediate,’’ and ‘‘modern’’ groups, and this rejected this position. He saw the origin of
grade-based practice has been followed, H. sapiens as a strictly African phenom-
explicitly and implicitly, in many subsequent enon, and regarded OH9 as morphologically

8. Table 1 Later African Hominidae, their archaeological associations and dates
460
Site Specimen Archaeology Date Method Selected references
Group 1 (H. erectus, H. ergaster, H. louisleakeyi, H. rhodesiensis)
Aın Maarouf
¨ Left femoral shaft Acheulian Early Middle Associated fauna Hublin, 1992;
(El Hajeb), Morocco Pleistocene Geraads et al., 1992
Berg Aukas, Namibia Femoral fragment None Undated Grine et al., 1995
Bodo, Ethiopia Adult cranium, Acheulian Mid to later Middle Associated fauna Conroy et al., 1978;
parietal, distal humerus Pleistocene, 350 ka Kalb et al., 1980, 1982a,b;
Asfaw, 1983; Clark et al.,
1984; Rightmire, 1996
40
cf. Oldowan 640 ka–550 ka Ar/39Ar, associated Clark et al., 1994
fauna
Buia, Danakil (Afar) Adult cranium, 2 None reported 1·0 Ma Paleomagnetism, Abbate et al., 1998
Depression, Eritrea incisors, pelvic associated fauna
fragments
Cave of Hearths, Mandible, radius Acheulian Early Late Associated fauna Cooke, 1962; Mason, 1962;
.   . . 
South Africa Pleistocene Mason et al., 1988; Tobias,
1971; Partridge, 1982;
Pearson Grine, 1997
End Middle Associated fauna Howell, 1978
Pleistocene
Eyasi, Tanzania Cranial fragments Sangoan 130 ka Extrapolation from Cooke, 1963; Mehlman,
representing 3–4 overlying 14C dates, 1984, 1987
individuals underlying 230Th/234U
dates, faunal
correlation

9. Table 1, Group 1 Continued
Site Specimen Archaeology Date Method Selected references
Kabwe (Broken Hill), Adult cranium (E686), ?Sangoan or MSA? 125 ka Associated fauna Woodward, 1921; Pycraft
Zambia (holotype of cranial, maxillary et al., 1928; Oakley, 1957;
H. rhodesiensis) dental and postcranial Clark, 1959; Clark et al.,
(humeral, pelvic, 1950, 1968; Klein, 1973,
femoral, tibial) remains 1994; Santa Luca, 1978;
of d3 individuals Partridge, 1982; Vrba,
1982; Stringer, 1986
110 ka Aspartic acid Bada et al., 1974
racemization on
hominid femoral
fragment EM 793
700–400 ka Associated fauna (cf. Klein, 1994; Rightmire,
Olduvai Beds III–IV) 1998
780 ka–1·33 Ma Associated fauna (cf. This paper, based upon
Olduvai Beds III–IV). Klein, 1973, 1994;
Correlation of top of Partridge, 1982; Hay, 1976;
Bed IV with Matuyama Walter et al., 1991, 1992;
   ’
Brunhes boundary Tamrat et al., 1995;
Kimbel, 1995; Delson
van Couvering, 2000
1·07–1·33 Ma Associated fauna (cf. This paper, based upon
Olduvai Beds III–IV). Klein, 1973, 1994;
Correlation of normal Partridge, 1982; Hay, 1976;
polarity paleomagnetic Walter et al., 1991, 1992;
zone at base of Masek Tamrat et al., 1995;
Beds with Jaramillo Kimbel, 1995; Delson
subchron van Couvering, 2000
461

10. Table 1, Group 1 Continued
462
Site Specimen Archaeology Date Method Selected references
Kapthurin (Baringo) Two adult mandibles, Undiagnostic 230–780 ka K/Ar, associated fauna, Leakey et al., 1969; van
Kenya (KNM-BK 67, 8518) paleomagnetism Noten, 1982; Howell, 1982;
postcranials (ulna, van Noten Wood, 1985;
talus, manus phalanges, Wood van Noten, 1986;
KNM-BK 63–66) Tallon, 1978; Dagley et al.,
1978; Rightmire, 1980;
Solan Day, 1992; Wood,
1992; Groves, 1998
500–550 ka Deino McBrearty, under
review
230
´
Kebibat (Rabat), Subadult calvaria, None 200 ka Th/234U Stearns Thurber, 1965
Morocco maxillary fragment,
mandible
300 ka–1·0 Ma Associated fauna Saban, 1975, 1977; Howell,
1978; Sausse, 1975b
40
Lainyamok, Kenya Femoral shaft, isolated Undiagnostic 390–330 ka Ar/39Ar Shipman et al., 1983; Potts
teeth et al., 1988; Potts Deino,
1995
.   . . 
Loyangalani Maxillary and None Late Middle/Early Associated fauna Twiesselmann, 1991
mandibular dentition Late Pleistocene
´
Melka Konture, Cranial fragments Acheulian Middle Pleistocene Associated fauna Chavaillon et al., 1974;
Ethiopia Howell, 1978; Chavaillon,
1982
Ndutu, Tanzania Adult cranium cf. Acheulian 500–600 ka AAR on associated Mturi, 1976; Rightmire,
mammalian bone 1980, 1983; Clarke, 1976,
1990

11. Table 1, Group 1 Continued
Site Specimen Archaeology Date Method Selected references
Ndutu, Tanzania 490–780 ka Upper limit: correlation This paper, based upon
continued of reversed polarity Tamrat et al., 1995
excursion in underlying
Norkilili Member of
Masek Beds with
Emperor subchron.
Lower limit: correlation
of top of Bed IV with
Matuyama Brunhes
boundary
370–990 ka Upper limit: 40Ar/39Ar This paper, based upon
on Kerimasi caldera, Hay, 1976, 1990; Leakey
probable source of et al., 1972; Leakey Hay,
Norkilili Member of 1982; Clarke, 1976, 1990;
Masek Beds. Lower Manega, 1993; Tamrat
limit: correlation of et al., 1995; Walter et al.,
normal polarity 1991, 1992; Kimbel, 1995
paleomagnetic zone at
base of Masek Beds
with Jaramillo
subchron
   ’
OH 9, LLK, Olduvai Adult partial cranium Developed Oldowan or 1·15 Ma (surface, top K/Ar, paleomagnetism Hay, 1963, 1973, 1976,
Gorge, Tanzania Acheulian of Bed II) 1990; Leakey, 1961, 1963;
(type of H. Leakey, 1971a,b; Leakey
louisleakeyi) Hay, 1982; Rightmire,
1979a, 1980, 1990, 1994;
Wood, 1994
0·7–1·0 Ma K/Ar, paleomagnetism Hay, 1971
40
1·33–1·48 Ma Ar/39Ar, Hay, 1976, 1990; Manega,
paleomagnetism 1993; Tamrat et al., 1995;
Walter et al., 1991, 1992;
Kimbel, 1995; White, 2000
463

12. Table 1, Group 1 Continued
464
Site Specimen Archaeology Date Method Selected references
OH 11, DK, Olduvai Palate, maxilla None; both ca. 400 ka (surface, K/Ar, paleomagnetism Rightmire, 1979a; Leakey
Gorge, Tanzania Acheulian MSA probably lower Ndutu inferred sedimentation Hay, 1982; Hay, 1994
elsewhere in Ndutu Beds) rates
Beds
ca. 490 ka Paleomagnetism; This paper, based upon
correlation of reversed Tamrat et al., 1995
polarity excursion in
underlying Norkilili
Member of Masek
Beds with Emperor
subchron.
40
ca. 370 ka Ar/39Ar on Kerimasi Hay, 1976, 1990; Manega,
caldera, probable 1993; Tamrat et al., 1995;
source of Norkilili Walter et al., 1991, 1992;
Member of Masek Kimbel, 1995; White, 2000;
Beds. Delson van Couvering,
2000
OH 12, VEK, Olduvai Palate, maxilla, cranial None; Acheulian 780–620 ka (upper Bed K/Ar, paleomagnetism, Rightmire, 1979a; Leakey
Gorge, Tanzania fragments elsewhere in Bed IV IV) inferred sedimentation Hay, 1982; Hay, 1994;
rates Leakey Roe, 1994
.   . . 
780 ka–1·2 Ma Paleomagnetism; This paper, based upon
correlation of top of Tamrat et al., 1995
Bed IV with Matuyama
Brunhes boundary
40
1·07–1·2 Ma Ar/39Ar Hay, 1976, 1990; Manega,
paleomagnetism; 1993; Tamrat et al., 1995;
correlation of normal Walter et al., 1991, 1992;
polarity paleomagnetic Kimbel, 1995; White, 2000;
zone at base of Masek Delson van Couvering,
Beds with Jaramillo 2000
subchron

13. Table 1, Group 1 Continued
Site Specimen Archaeology Date Method Selected references
OH 22, VEK/MNK, Partial mandible None; Acheulian 800–600 ka (surface, K/Ar, paleomagnetism, Day, 1986; Leakey Hay,
Olduvai Gorge, elsewhere in Bed IV Bed IV) inferred sedimentation 1982; Leakey Roe, 1994;
Tanzania and overlying Ndutu rates Rightmire, 1979a, 1990;
Beds Hay, 1994
780 ka–1·2 Ma Paleomagnetism; This paper, based upon
correlation of top of Tamrat et al., 1995
Bed IV with Matuyama
Brunhes boundary
40
1·07–1·2 Ma Ar/39Ar, Hay, 1976, 1990; Manega,
paleomagnetism; 1993; Tamrat et al., 1995;
correlation of normal Walter et al., 1991, 1992;
polarity paleomagnetic Kimbel, 1995; White, 2000;
zone at base of Masek Delson van Couvering,
Beds with Jaramillo 2000
subchron
OH 23, FLK, Olduvai Mandibular fragment Acheulian 400–600 ka (In situ, K/Ar, paleomagnetism, Day, 1986; Rightmire,
Gorge, Tanzania Masek Beds) inferred sedimentation 1990; Leakey Roe, 1994;
rates Hay, 1994
   ’
490–780 ka Paleomagnetism. This paper, based upon
Upper limit: correlation Tamrat et al., 1995
of reversed polarity
excursion in underlying
Norkilili Member of
Masek Beds with
Emperor subchron.
Lower limit: correlation
of top of Bed IV with
Matuyama Brunhes
boundary
465

14. Table 1, Group 1 Continued
Site Specimen Archaeology Date Method Selected references
466
OH 23, continued 990–370 ka Upper limit: 40Ar/39Ar Hay, 1976, 1990; Manega
on Kerimasi caldera, 1993; Tamrat et al., 1995;
probable source of Walter et al., 1991, 1992;
Norkilili Member of Kimbel, 1995; White, 2000;
Masek Beds. Lower Delson van Couvering,
limit: correlation of 2000
normal polarity
paleomagnetic zone at
base of Masek Beds
with Jaramillo
subchron
OH 28, WK, Olduvai Left innominate, femur Acheulian 780–620 ka (upper Bed K/Ar, paleomagnetism, Day, 1971, 1986; Leakey,
Gorge, Tanzania IV) inferred sedimentation 1971a; Rightmire, 1979a;
rates Leakey Hay, 1982; Hay,
1994
780 ka–1·2 Ma Paleomagnetism; This paper, based upon
correlation of top of Tamrat et al., 1995
Bed IV with Matuyama
Brunhes boundary
40
1·2–1·07 Ma Ar/39Ar, Hay, 1976, 1990; Manega,
paleomagnetism; 1993; Tamrat et al., 1995;
.   . . 
correlation of normal Walter et al., 1991, 1992;
polarity paleomagnetic Kimbel, 1995; White, 2000;
zone at base of Masek Delson van Couvering,
Beds with Jaramillo 2000
subchron
OH 34, JK, Olduvai Femur and partial tibia Acheulian 0·8–1·1 Ma (Bed III) K/Ar, paleomagnetism Day, 1971; Leakey Roe,
Gorge, Tanzania 1994; Hay, 1990, 1994
40
1·2–1·33 Ma Ar/39Ar, This paper, based upon
paleomagnetism Hay, 1976, 1990; Tamrat
et al., 1995; Walter et al.,
1991, 1992; Kimbel, 1995;
White, 2000

15. Table 1, Group 1 Continued
Site Specimen Archaeology Date Method Selected references
OH 51, GTC, Mandibular fragment None; Acheulian 0·8–1·1 Ma (Bed III) K/Ar, paleomagnetism Leakey Roe, 1994;
Olduvai Gorge, elsewhere in Bed III Rightmire, 1990; Hay,
Tanzania 1990, 1994
40
1·2–1·33 Ma Ar/39Ar, This paper, based upon
paleomagnetism Hay, 1976, 1990; Tamrat
et al., 1995; Walter et al.,
1991, 1992; Kimbel, 1995;
White, 2000
Saldanha (Hopefield Adult calvaria, Acheulian ca. 130–780 ka (Middle Associated fauna Drennan, 1953; Singer,
Elandsfontein), South mandibular fragment Pleistocene) 1954; Cooke, 1963;
Africa Partridge, 1982; Deacon,
1988
600–800 ka Associated fauna, cf Klein, 1973; Singer
Bed IV Olduvai Gorge Wymer, 1968; Leakey
Hay, 1982; Hay, 1994;
Leakey Roe, 1994
   ’
500–200 ka Associated fauna Klein, 1988
700–400 ka Associated fauna Klein Cruz-Uribe, 1991;
Klein, 1994; Rightmire,
1998
780 ka–1·2 Ma Associated fauna (cf This paper, based upon
Olduvai Bed IV). Gentry, 1978; Klein Cruz
Correlation of top of Uribe, 1991; White, 2000;
Bed IV with Matuyama Tamrat et al., 1995; Walter
Brunhes boundary et al., 1991, 1992; Kimbel,
1995
467

16. Table 1, Group 1 Continued
468
Site Specimen Archaeology Date Method Selected references
Saldanha, continued 1·07–1·33 Associated fauna (cf This paper, based upon
Olduvai Bed IV). Gentry, 1978; Klein Cruz
Correlation of normal Uribe, 1991; White, 1999;
polarity paleomagnetic Tamrat et al., 1995; Walter
zone at base of Masek et al., 1991, 1992; Kimbel,
Beds with Jaramillo 1995
subchron
´
Sale, Morocco Adult calvaria, cranial None 300 ka–1·0 Ma Associated fauna Jaeger, 1973, 1975; Howell,
fragments, endocast 1978; Hublin, 1985, 1991,
1994; Dean et al., 1993
389–455 ka LU ESR on associated Hublin, 1991
bovid tooth enamel
Sidi Abderrahman Partial mandible Acheulian Middle Pleistocene Geomorphology, Arambourg Biberson,
(Casablanca) Morocco (‘‘Tensiftian’’=‘‘Riss’’) associated fauna 1956; Biberson, 1963;
Howell, 1960, 1978
´
Tighenif (formerly Three mandibles, Acheulian Middle Pleistocene Associated fauna Arambourg, 1955;
Ternifine), Oran, parietal fragment Arambourg Hoffsteter,
Algeria 1963; Howell, 1960, 1978;
Balout et al., 1967; Tobias,
.   . . 
1968; Schwartz
Tattersall, 2000
0·6–1·0 Ma Associated fauna Geraads, 1981; Jaeger, 1981
700 ka Associated fauna, Geraads et al., 1986
paleomagnetism
Thomas 1 Quarry, Subadult partial None Middle Pleistocene Associated fauna, Ennouchi, 1969a, 1970;
Morocco mandible, cranial and (‘‘Amirian’’) geomorphology Sausse, 1975b; Oakley et al.,
maxillary fragments 1977; Howell, 1960, 1978;
Brauer, 1984a,b; Dean
¨
et al., 1993

17. Table 1, Group 1 Continued
Site Specimen Archaeology Date Method Selected references
´
Wadi Dagadle, Maxilla, partial None 250 ka TL on basalt, de Bonis et al., 1984, 1988
Djibouti dentition associated fauna
Group 2 (H. helmei or H. sapiens)
Eliye Springs, West Adult cranium None None Surface find Brauer Leakey, 1986
¨
Turkana, Kenya ES11693
14
Florisbad, South Adult cranium ?MSA? Infinite AAR, C Vogel Beaumont, 1972;
Africa Bada et al., 1973
230
100 ka Th/234U, peat I, Dreyer, 1935, 1936;
associated fauna Rightmire, 1978a; Clarke,
1985; Kuman Clarke,
1986; Brink, 1988; Butzer,
1988a; Kuman et al., 2000
260 ka ESR, direct assay on Grun et al., 1996
¨
hominid tooth
   ’
Guomde, Chari Fm, Adult cranium None 270–300 ka U-series, direct assays Brauer et al., 1992, 1997;
¨
Ileret, Kenya (KNM-ER 3884), on hominid cranium Feibel et al., 1989
femur (KNM-ER 999) femur
14
Haua Fteah, Libya Two young adult MSA (‘‘Levalloiso- Infinite C on burnt bone McBurney, 1961, 1967;
mandibular fragments Mousterian’’) Tobias, 1967; Rak,
1998:364
130 Associated artefacts This paper, based upon
(Generalized MSA is Debenath, 1994; Wendorf
´
stratified under Aterian et al., 1987, 1993a
elsewhere in Sahara)
469

18. 470
Table 1, Group 2 Continued
Site Specimen Archaeology Date Method Selected references
Haua Fteah, Libya 90 ka Associated artefacts. This paper, based upon
continued TL OSL reported Martini et al., 1996;
for in situ Aterian Cremaschi et al., 1998
material in Libya;
generalized MSA is
stratified under Aterian
elsewhere in Sahara
127–40 ka Associated artefacts Klein, 1999
14
Jebel Irhoud, Adult cranium (JI1), MSA (‘‘Levalloiso- Infinite C Ennouchi, 1966
Morocco adult calvaria (JI2), Mousterian’’)
infant mandible (JI3),
fragmentary
postcranials (JI4)
90–125 ka (EU), 105 Extrapolation from Ennouchi, 1962, 1963,
190 ka (LU) ESR dates on 1968, 1969a,b; Howell,
mammalian teeth 1978; Hublin, 1985, 1991,
overlying in situ 1993; Hublin et al., 1987;
hominid specimen JI4 Grun Stringer, 1991
¨
.   . . 
Mugharet el Aiya, Juvenile maxillary MSA, presumed Late Middle/Early Associated fauna Biberson, 1961, 1963;
Morocco fragment with partial Aterian (not in situ) Late Pleistocene Debenath, 1980; Debenath
´ ´
dentition, adult tooth (‘‘Ouljian-Soltanian’’) et al., 1982, 1986; Amani
Geraads, 1993; Hublin,
1993
60–90 ka Associated Aterian This paper, based upon
artifacts; TL OSL Debenath, 1994; Martini
´
reported for in situ et al., 1996; Cremaschi
Aterian material in et al., 1998
Libya

19. Table 1, Group 2 Continued
Site Specimen Archaeology Date Method Selected references
Ngaloba (Laetoli Adult cranium MSA 120 ka Correlation with Hay, 1976; Leakey Hay,
Hominid 18) marker tuff in Lower 1982; Magori Day, 1983;
Tanzania Ndutu Beds at Olduvai Day et al., 1980
Gorge bracketed by
14
C K/Ar dates
230
130–108 ka Th/234U on Hay, 1987
associated mammalian
bone
100–200 ka Isoleucine Bada, 1987
epiminerization of
associated mammalian
bone
200 ka AAR on ostrich Manega, 1995
eggshell, correlation
with units dated by
40
Ar/39Ar
40
200–370 ka Ar/39Ar, on Kerimasi This paper, based on Hay,
   ’
caldera, probable 1976, 1990; Manega, 1993;
source of Norkilili Tamra et al., 1995; Walter
Member of Masek et al., 1991, 1992; Kimbel,
Beds. 1995
40
200–490 ka Ar/39Ar, This paper, based on Hay,
paleomagnetism; 1976, 1990; Manega, 1993;
correlation of reversed Tamrat et al., 1995; Walter
polarity excursion in et al., 1991, 1992; Kimbel,
underlying Norkilili 1995; Delson van
Member of Masek Couvering, 2000
Beds with Emperor
subchron.
471

20. 472
Table 1, Group 2 Continued
Site Specimen Archaeology Date Method Selected references
14
Omo II, Site PHS, Adult calvaria ?MSA? 39 ka C on Etheria shell in Day, 1969, 1972; Butzer,
Kibish Formation, overyling unit 1969; Butzer et al., 1969;
Ethiopia Merrick et al., 1973; Day
Stringer, 1982
230
130 ka Th/234U on Etheria Day, 1969, 1972; Butzer,
shell in overlying unit 1969; Butzer et al., 1969;
Merrick et al., 1973; Day
Stringer, 1982, 1991
Porc Epic (Dire
u Mandibular fragment MSA Late Pleistocene Associated fauna Breuil et al., 1951; Vallois,
Dawa), Ethiopia 1951; Clark, 1954, 1982;
Brauer, 1984a; Howell,
¨
1978
60–77 ka Obsidian hydration Clark et al., 1984; Clark,
1988
Singa, Sudan Calvaria ?MSA? Early Late Pleistocene Associated fauna Bate, 1951; Stringer, 1979;
Brauer, 1984a,b; Stringer
¨
.   . . 
et al., 1985; Clark, 1988
82–112 (EU), 133–187 ESR on associated Grun Stringer, 1991
¨
(LU) mammalian teeth
190–130 ka U-series on calcrete McDermot et al., 1996
enclosing skull and
associated mammalian
teeth

21. Table 1 Continued
Site Specimen Archaeology Date Method Selected references
Group 3 (H. sapiens)
Border Cave, South Adult calvaria (BC1), 2 MSA 90–115 ka Geomorphology, Cooke et al., 1945; Wells,
Africa adult mandibles (BC 2 associated fauna, 1950, 1959; de Villiers,
5), infant partial extrapolation from 1973, 1978; Protsch, 1975;
skeleton (BC3), adult overlying 14C dates, Beaumont et al., 1978;
postcranial fragments amino acid racimzation Butzer et al., 1978;
Rightmire, 1979b;
Beaumont, 1980; Grun et al.,
¨
1990; Grun Stringer,
¨
1991; Morris, 1992b; Miller
et al., 1993, 1999; Pfeiffer
Zehr, 1996; Pearson
Grine, 1996 (but see Sillen
Morris, 1996)
90–50 ka AAR, associated fauna, Klein, 1999
geomorphology,
extrapolation from
overlying 14C dates
14
Dar-es-Soltan, Cranial, maxillary MSA (Aterian) Infinite C McBurney, 1961;
   ’
Morocco mandibular fragments Ferembach, 1976b; Howell,
of 2 individuals, adult 1978; Brauer Rimbach,
¨
subadult 1990; Wendorf et al., 1990;
Brauer, 1992; Debenath,
¨ ´
1980, 1994; Debenath
´
et al., 1982, 1986; Hublin,
1993
60–90 ka Associated Aterian This paper, based upon
artifacts; TL OSL Debenath, 1994; Martini
´
for in situ Aterian et al., 1996; Cremaschi
material in Libya et al., 1998
473

22. 474
Table 1, Group 3 Continued
Site Specimen Archaeology Date Method Selected references
Die Kelders, South 24 teeth, mandibular MSA 40 ka, 60 ka (EU), ESR Grine Klein, 1985; Grine
Africa fragment, 2 manual 80 ka (LU) et al., 1991; Avery et al.,
phalanges 1997
71–45 ka ESR, associated fauna, Klein, 1999
geologic context
60–80 ka ESR, TL, OSL, IRSL Grine, 2000; Schwarcz
Rink, 2000; Feathers
Bush, 2000
14
Equus Cave, South Mandibular left corpus MSA 93–44 ka U-series on tufa, C Grine Klein, 1985; Klein
Africa (hyena layer) fragment with 2 on Mn patina et al., 1991
molars; additional
isolated adult teeth
14
71–27 ka C, associated fauna, Klein, 1999
geologic context
.   . . 
Hoedjies Punt, South Cranial and postcranial MSA 70 ka U-series Volman, 1978; Berger
Africa fragments, isolated Parkington, 1995
teeth
300–71 ka U-series, associated Klein, 1999
fauna, geologic context
Kabua, Kenya Cranial mandibular Uncertain ?Late Pleistocene Geologic context Whitworth, 1966;
fragments of 2 Rightmire, 1975
individuals