Current events

Richard Potts Olorgesailie: new excavations and findings

in Early and Middle Pleistocene contexts,
Department ofAnthropology,
National Museum qf Natural southern Kenya rift valley
History, Smithscwiarz Institution,
Washington, DC 20560, U.S.A. and
National Museum of Kenya,
P.O. Box 4ossS, Nairobi, Kenya

Journal of Human Evolution (1989) l&477-484

Introduction

Since 1985, new excavations and surveys have been carried out at this rift valley locality by
a Smithsonian Institution research team in conjunction with the National Museums of
Kenya, Nairobi. Olorgesailie (1’35’s 36”27’E) is a Pleistocene lake basin that contains
sediments representing lacustrine, lake margin, deltaic, and fluvial environments during a
period of regional volcanic activity (Figure 1). Acheulean handaxes, a dominant
component of hominid lithic technology for over a million years throughout much of the
Old World (l~l-02 Ma), are extremely abundant at Olorgesailie. Because ofthis, previous
excavations at this locality from 1942-1945 by L. S. B. and M. D. Leakey and from
1962-1965 by G. L. Isaac focused on the recovery of stone handaxes. Leakey’s goal was to
learn about “Acheulean culture” (Leakey, 1952). Isaac (1968a,6, 1977) carried out a

KENYA
/ I

Figure 1. Location of the Olorgesailie basin and its relationship to Lainyamok and other Pleistocene
localities in East Africa.

0047-2484/89/050477 + 08 $03.00/O 0 1989 Academic Press Limited

478 CURRENT EVENTS

detailed study of stylistic variation in the handaxe assemblages, and tended to view
different repeated styles of tool production as reflecting divisions of a population into
separate social groups and the maintenance of distinctive habits of tool manufacture
(Isaac, 1977: 96,211). The majority of sites previously excavated at Olorgesailie have been
interpreted as occupation sites, or stable home bases, of hunter-gatherers. One site
(DE/89B) preserved numerous bones of extinct gelada monkeys (Theropithecus oswaldi) and
is thought to represent systematic hunting and butchery by hominids (Isaac, 1977;
Shipman et al., 1981).
However, the problem with the interpretation of previous excavations at Olorgesailie is
that the sites were confined primarily to a fluvial facies identified by silts, sands, gravels,
pebble conglomerates, and channel scour features (Isaac, 1977; Owen & Renaut, 1981).
Such contexts do not favor direct translation of archeological patterning into hominid
activities due to the potential for rapidly moving water within channels to move bones and
stone artifacts and to mix the materials from sites created at widely different times and
places (Binford, 1977; Hanson, 1980; Koch, 1986). Taphonomic analyses have illustrated
repeatedly the need to evaluate the effects ofwater action, hominid and carnivore activities,
and trampling and other forms of disturbance in order to derive valid inferences about
paleontological and archeological bone assemblages. New excavations devoted to this type
of analysis were necessary at Olorgesailie, which based on previous excavations has
become a textbook example of evidence presumed to pertain directly to Middle Pleistocene
hominid activities.

Geology and dating

The Olorgesailie Formation is divided into 14 members distinguished on the basis of

lithology and stratigraphic position (Isaac, 1978). A map of the members within the basin
was made by Shackleton ( 1978)) and his recognition of the members has been supported by
detailed stratigraphic studies (Isaac, 1977, 1978) and tuff chemistry (Bye, 1984; Bye et al.,
1987).
An age of 0.4-0.5 Ma has traditionally been accepted for Olorgesailie. This was based on
two conventional K/Ar dates (0.42 and 0.48 Ma) out of five samples analyzed during the
early 1960s (Evernden & Curtis, 1965; Isaac, 1977). Thus, Olorgesailie has been
considered to span a relatively brief period during the middle part of the Middle
Pleistocene. In fact, the fauna and stone tools derived from this locality have long been
assumed to characterise the Middle Pleistocene in sub-Saharan Africa. A recent study,
however, concludes that the upper part of the formation (Members lo- 14) ranges between
O-74 and O-6 Ma, based on two dates from Member 10 of the Olorgesailie Formation and a
correlation of tuff chemistry to a volcanic cone (01 Doinyo Nyokie) believed to be that old
(Bye et al., 1987).
During the past several months the Smithsonian research project in collaboration with
the Berkeley Geochronology Laboratory has obtained a series of +‘JAr-sgAr laser-fusion
dates which shows that the Olorgesailie Formation ranges from approximately 0.99 Ma
(Member 1) to 0.49 Ma (Member 14); dates of approximately 0.22 and 0.05 Ma have also
been obtained from newly recognised parts of the Olorgesailie sequence which overlie the
previously defined members of the formation (Deino & Potts, in preparation). Analyses
aimed at identifying paleomagnetic reversals in the sequence are in progress.
Based on the *OAr-s9Ar results, Olorgesailie spans the Early through Late Pleistocene
and represents the most securely dated artifact- and fossil-rich sequence spanning a length
 CURRENT iW@TS 479

and represents the most securely dated artifact- and fossil-rich sequence spanning a
lengthy period immediately preceding and including the transition from Homo erectus to
Homo sapiens in sub-Saharan Africa. Although no hominid fossils have yet been uncovered
from this locality, abundant stone artifacts and associated fauna have now been found in
fine-grain, non-fluvial sediments. This means that Olorgesailie preserves the kind of
evidence present in Bed I Olduvai and Koobi Fora on which interpretations and recent
reanalyses of Plio-Pleistocene hominid behavior and paleocology have been based (e.g.,
Bunn et al., 1980; Bunn & Kroll, 1986; Isaac, 1984; Leakey, 1971; Potts, 1988; Toth &
Schick, 1986).

New excavations and finds

Given the primarily channel or fluvial-related contexts of previous excavations, the aim of
an initial survey by the Smithsonian team in 1985 was to let the distribution of tine-grain
sediments rather than the presence of handaxes indicate the best places for future
excavation. Sixteen new sites or fossil-rich areas were found in the Olorgesailie basin. From
1986-1988, excavations have focused on an outcrop of line silts in a lake margin facies that
had not been previously recognised at Olorgesailie. This outcrop (the vicinity of Leakey’s
Friday Beds) is near the top of Member 1, the oldest unit in the formation. The main layer
(named layer 4) that bears fossils and lithic artifacts is primarily a yellow-brown paleosol
composed of tuffaceous silts and clays with abundant root marks and burrows. Situated
between two thick units of diatomite which were deposited in quiet lake waters, the silts of
the main fossiliferous layer were deposited during a period of lake regression. Minor
portions oflayer 4 that were not pedogenically modified indicate that a grey-colored, sandy
to silty tuff was originally deposited in a floodplain and shallow channel system that
traversed an exposed area of lake flats. Evidence of soil formation implies that a stable land
surface was then formed permitting vegetation and other soil processes to alter the color
and texture of the tuff. The thickness of the layer 4 paleosol varies laterally from ca. 10 to
120 cm; the variation in thickness appears to have permitted differences in the potential for
bone and artifact burial and preservation. The vast majority of hominid artifacts and
animal bones is associated with the paleosol, suggesting that their deposition and burial
were contemporaneous with paleosol formation.
The paleosol outcrop is exposed continuously by erosion over a distance of 4-5 km,
which represents a winding transect through the ancient lake margin. This erosional
pattern and the presence of very little overburden 2 to 5 m back from the erosion slope
allow us to sample the distribution of artifacts and animal remains originally deposited on
a land surface ca. 0.99 Ma. By making this transect through the paleosol the target of
sampling, the spatial range of hominid activities and habitats under investigation is greatly
expanded when compared to traditional approaches that focus on small exposures of
dense, isolated clusters of artifacts and bones.
In more than 20 excavation areas, approximately 110 m2 of this paleosol have been
exposed so far, representing a small sampling of 2 km of the erosional transect. Excavations
spaced over this 2 km transect indicate that the lateral distribution of stone artifacts and
fossil bones within the paleosol is virtually continuous, but with considerable variation in
the degree of concentration of these objects. Based on the spatial clustering of artifacts and
cutmarked bones, two main areas of hominid activity have been detected so far over the
lateral extent of the paleosol.
480 CURRENT EVENTS

One excavation (Site 15), 23 m2 in area, consists of most of the skeleton of an elephant,
E1epha.s recki, spatially associated with stone artifacts. With the exception of five or six small
pieces identifiable to other species, the fossilised bones of the elephant are isolated from the
remains ofother animals. The artifacts (N> 350) occur amongst the ribs, mandible, and in
small clusters around other elements in the elephant carcass scatter. The artifacts consist
mainly of sharp stone flakes in mint condition, some possessing areas of chipping that
appears to be due to utilisation. Although investigation of this site is still in progress, it is
currently interpreted to be a locus of butchery of an elephant that died at the site. Virtually
every other Early Pleistocene site purported to be a large animal butchery area has been
reinterpreted taphonomically in recent years, generally due to the presence of numerous
bones of other animals (e.g., Crader, 1983; Potts, 1988). Hence, the spatially exact and
virtually exclusive association between the elephant skeleton and stone artifacts is an
exceptional find that may prove to be an important example of hominid utilisation of a
large mammal carcass.
A second area, which based on surface finds spans several hundred mz, has been exposed
in six excavation trenches totalling approximately 28 m2. The main trench (referred to as
Site 2) exhibits a relatively dense accumulation of fossils (mean of 18.7 specimens per ms)
intimately associated with a relatively dense concentration of artifacts (mean of 15.7
specimens per m2). The spatial density of remains is approximately two to five times the
concentration of materials in other areas of the paleosol excavated so far (with the
exception of Site 15). Three of the excavation trenches in the vicinity of Site 2 have yielded
bones with stone tool cut marks. Although this site was undoubtedly an area of repeated
hominid activity, it has not yet been determined whether it represents the cumulative effect
of numerous visits by hominids to a favored foraging area, or repeated transport of bones
and artifacts by hominids to either a home base or a carcass processing site. The distinction
between these interpretations at other localities (e.g., Olduvai) has contributed to
important debates about the behavior and evolution of early hominids (e.g., Isaac, 1984;
Potts, 1984, 1988; Binford, 1981).
Another area we began to excavate in 1988 (Site 102)) located about 10 meters from Site
2, preserves at least three of the most complete skeletons of mammalian carnivores known
from the African Pleistocene. Two are hyena (Crocuta) and one is a canid. The complete
skeletons of these animals are probably present (complete crania to sesamoid bones are
preserved). The bones occur in a complicated set of depressions in the paleosol that
represent burrows possibly used by these animals prior to death. Two meters to the north,
excavation of another depression in the paleosol (Site 104) has begun to uncover an
accumulation of ungulate bones, also collected originally in a burrow. The bones include
well-preserved postcranial remains and a complete cranium of the extinct zebra Equus
oldowayensis. Due to the intricate nature of excavating the clusters of bones at these two
sites, both were reburied at the conclusion of the 1988 season and will be re-excavated in
1989.

Discussion
The new excavations at Olorgesailie are intended to explore an ongoing set of questions
concerning early hominid foraging, social behavior and ecological history. I will confine
comments to three areas: hominid land use, ecological overlap with carnivores and
influences on the spatial distribution of hominid activities.
 CURRENT EVl%NTS 481

Hominid land use

Variation in the presence and types of stone artifacts in different sedimentary
environments may reflect differential use ofhabitats by hominids. The artifact assemblages
excavated from the Member 1 paleosol consist almost entirely of sharp flakes, flake
scrapers and Oldowan-type cores. In contrast to assemblages in the channel contexts of
previous excavations, handaxes constitute a minor component in the lake margin zone
(< 5% of the total number of flaked pieces vs. 62% at channel-related sites). Since no
handaxes comparable to the size of those at the channel sites have been found in the lake
margin paleosol, other factors besides taphonomic size sorting must have helped to
produce this biased distribution of artifact types. The association ofhandaxes with channel
facies and of handaxe-poor assemblages with floodplain or lake margin zones has been
documented at other localities (e.g., Hansen & Keller, 1971; Hay, 1976). Isaac (1977) also
noted such a correlation based on prior excavations at Olorgesailie. These associations
appear to reflect a spatial patterning of tool use and artifact discard that persisted over
perhaps a million years during the Pleistocene in East Africa. Its significance in terms of
hominid foraging and land use is not yet known.
At Olorgesailie, however, this differential facies distribution of artifacts appears to be
related to the function of handaxes and other bifaces-a long-standing, unresolved
question about this and other Acheulean localities. Portions of the edges of stone flakes at
the elephant site, for example, exhibit fine (non-retouch) chipping, possibly from use
during butchery of the elephant. During excavation, the flakes were also observed to
possess complicated striking platforms and intersecting scars on their dorsal surfaces. This
indicates that these flakes were derived from large bifacial cores. At Olorgesailie, the term
“large bifacial cores” can only mean bifaces such as handaxes. One function of handaxes,
therefore, may have been as valuable “blade dispensers”, carried but seldom discarded
while foraging, whereas the used and unused flakes derived from them were left behind.
The abundance of handaxes in the channel facies would then reflect proximity to their
place of manufacture, a focus upon other activities that required hominids to deposit the
handaxes, or the use of handaxes as implements for some other activity in this zone of the
landscape.

Ecological overlap with carnivores

Of approximately 22,000 fossil specimens recovered in the Leakey and Isaac excavations,
only three specimens and a minimum number of two individuals represent carnivores
(Koch, 1986). C arnivore fossils are considerably more abundant in the lake margin sample
uncovered from the paleosol during the past three field seasons. However, in places where
animal bones have been found associated with stone artifacts, tooth marks made by
carnivores have not yet been found during field excavation or preliminary examination of
bones in the laboratory.
At Lainyamok, a Middle Pleistocene locality 40 km southwest of Olorgesailie (Figure l),
animal bones appear to have been modified and concentrated exclusively by carnivores;
there is no evidence of hominid involvement in processing carcasses or exploiting game at
this locality (Potts et al., 1988). Coupled with initial study of the Olorgesailie finds,
hominids and carnivores appear not to have modified bones in the exact same places of the
landscape by the mid-Pleistocene in the southern Kenyan rift (after 1-O Ma), which
contrasts considerably with evidence from Bed I Olduvai, 1.85-1.70 Ma. At the Olduvai
artifact sites, bone damage by hominids and carnivores occur at the same sites and focus
482 CURRENT EVENTS

upon the same species, body parts, and (rarely) on the same bone specimens (Potts, 1984,
1988; Shipman 1986a,b).
I have suggested previously (Potts, 1984, 1988) that predator avoidance and a reduction
in the tight spatial overlap between large carnivores and hominids were necessary for the
development of safe, terrestrial social refuges where hominids also processed large
mammals for food. Such social and foraging foci would resemble home bases, or campsites,
of modern human hunter-gatherers. Initial comparisons among Olorgesailie, Lainyamok
and Olduvai suggest that these conditions were met in the southern Kenya rift by the
mid-Pleistocene. However, the presence of carnivores in the Member 1 lake margin zone at
Olorgesailie parallels that in Bed I Olduvai. Although exact spatial overlap in the use of
carcasses has not yet been found, evidence of an area of carnivore activity (Sites 102, 104)
in close proximity to an area of hominid activity (Site 2) provides an especially intriguing
complication. Continuing study of the spatial relationships between carnivore and
hominid activities in the lake margin paleosol will help to test and to reline these ideas
about ecological overlap.

Influences on the spatial distribuition of hominid activities

In Member 1, diatom species indicate the presence of a lake very low in salinity and
alkalinity (Figure 2), and the artifacts of Members 7, 10 and 11 are largely associated with
ephemeral stream channels that must have contained fresh water on at least a seasonal
basis (Isaac 1977; Owen & Renaut, 1981). The influence of available fresh water on the
distribution of hominid activities in the southern Kenya rift is further suggested at
Lainyamok, where hominid activity is poorly represented, and C and 0 isotopic values

TOTAL
MEMBER ALKALINITY SALINITY . LAKE DEPTH

14

13

12

11 \‘

9-10 \‘

u
3-8

I\\
I_
U 10 50 200 0 2 30 Shallow Deep

meq/l Q/l

Figure 2. Estimates ofalkalinity, salinity, and depth (area1 extent) ofthe paleo-lake at Olorgesailie (from
Owen & Renaut, 1981), and the statigraphic position of stone artifact occurrences in the Olorgesailie
basin (shaded bars). Lake curves are based on the presence ofdiatom species that are sensitive indicators
of lake water chemistry and size, and on the distribution of lithofacies in the basin.
 CURRENT EVENTS 483

indicate the presence of an extremely saline and alkaline lake (Potts et al. 1988; E. Ingall,
personal communication). The lake curves for Olorgesailie also indicate that small lake
size (and shallow depth) correlates with the concentration of hominid activities within the
current area of erosional exposure in this sedimentary basin (Figure 2). Based on these
preliminary comparisons, the spatial distribution of hominid activities appears to have
been influenced by factors other than just the availability of meat-eating opportunities,
which apparently did exist at Lainyamok based on the collecting of ungulate bones by
carnivores. Abiotic factors such as the physical and chemical characteristics of water
bodies seem to have also played a role.
On a smaller scale, the 4-5 km long, fossil- and artifact-rich paleosol in Member 1 offers
an excellent opportunity to sample by excavation the distribution ofhominid activities and
other taphonomic agents over a paleolandscape. Areas of hominid activity can be
examined in relation to the original land surface topography (corrected for geologic dip),
general vegetation types (indicated by variations in root markings-e.g., swamp VS. grass),
and the location of diverse stone raw material sources within the Olorgesailie basin. This
paleo-landscape approach represents an improvement over previous research strategies
oriented toward extremely limited sampling of the spatial distribution, and probably the
behavioral range, of hominid activities during the Early and Middle Pleistocene of East
Africa.

Acknowledgements

I thank the Offtce of the President, Republic of Kenya, and R. E. Leakey and the National
Museums of Kenya for permission and support in conducting the field and laboratory
studies. Invaluable support in the field has been given by J. Muteti Nume, who leads the
Olorgesailie excavation crew, Thalassa W. Skinner and Anne Cooksey. The research
group includes A. K. Behrensmeyer and W. G. Melson (Smithsonian), A. Deino (Berkeley
Geochronology Laboratory), E. Ingall (Yale University), and D. Clark (Scripps
Institution). The figures were drawn by Jennifer Clark. Funding was provided by the
Smithsonian Institution’s Scholarly Studies Program.

References

Binford, L. R. (1977). Olorgesailie deserves more than the usual book review,J. Anthrop. Res. 33, 493-502.
Binford, L. R. (1981). Bones: Ancient Men and Modem Myths. Orlando: Academic Press.
Bunn, H. T. & Kroll, E. M. (1986). Systematic butchery by Plio/Pleistocene hominids at Olduvai Gorge,
Tanzania. Curr. Anthrop. 27, 431-452.
Bunn, H., Harris, J. W. K., Isaac, G., Kaufulu, Z., Kroll, E., Schick, K., Toth, N. & Behrensmeyer, A. K. (1980).
FxJj 50: An Early Pleistocene site in northern Kenya. Wld Arch. 12, 109136.
Bye, B. A. (1984). Volcanic statigraphy and red beds of the Olorgesailie Formation, southern Kenya. MS.
Thesis, University of Utah.
Bye, B. A., Brown, F., Cerling, T. & McDougall, I. (1987). Increased age estimate for the Lower Paleolithic
hominid site at Olorgesailie, Kenya, Nature, Land. 329, 237-239.
Crader, D. C. (1983). Recent single-carcass bone scatters and the problem of “butchery” sites in the
archaeological record. In (J. Glutton-Brock & C. Grigson, Eds) Animalsandrlrchaeology: I. Hunters and TheirPrey,
pp. 107-141. BAR International Series 163.
Evernden, J. F. & Curtis, G. H. (1965). The potassium-argon dating of late Cenozoic rocks in East Africa and
Italy, Curr. Anthrop. 6, 343-385.
Hansen, C. L. & Keller, C. M. (1971). Environment and activity patterning at Isimila Korongo, Iringa District,
Tanzania: A preliminary report, Am. Anthrop. 73, 1201-1211.
Hanson, C. B. (1980). Fluvial taphonomic processes: Models and experiments. In (A. K. Behrensmeyer & A.
Hill, Eds) Fossils in the Making, pp. 156181. Chicago: Chicago University Press.
484 CURRENT EVENTS

Hay, R. L. (1976). Geology ofthe Olduvai Gorge. Berkeley: University of California Press.
Isaac, G. L. (196&r). Traces of Pleistocene hunters: An East African example. In (R. Lee & I. DeVore, Eds) Man
the Hunter, pp. 253-261. Chicago: Aldine-Atherton.
Isaac, G. L. (19683). The Acheulean site complex at Olorgesailie, Kenya: A contribution to the interpretation of
Middle Pleistocene culture in East Africa. Ph.D. Dissertation, Cambridge University.
Isaac, G. L. (1977). Olorgesailie. Chicago: University of Chicago Press.
Isaac, G. L. (1978). The Olorgesailie Formation. In (W. W. Bishop, Ed.) Geological Backsround to Fossil Man,
pp. 173-206. Edinburgh: Scottish Academic Press.
Isaac, G. L. (1984). The archaeology of human origins. In (F. Wendorf & .4. E. Close. Eds) Advances in World
Archaeology, Vol. 3, pp. l-87. New York: Academic Press.
Koch, C. (1986). The vertebrate taphonomy and palaeoecology of the Olorgesailie Formation. Ph.D.
Dissertation, University of Toronto.
Leakey, L. S. B. (1952). The Olorgesailie prehistoric site. In (L. S. B. Leakey & S. Cole, Eds) Proc. First
Pan-A&can Gong. Preh., 1947, Nairobi, p. 209.
Leakey, M. D. (1971). Olduoai Gorge, Vol. 3. Cambridge: Cambridge University Press.
Owen, R. B. & Renaut, R. (1981). Paleoenvironments and sedimentology of the Middle Pleistocene Olorgesailie
Formation, southern Kenya rift valley. In (J. Coetzee & E. M. van Zinderen Bakker, Eds) Palaeoccology ofAfrica,
pp. 147-174. Rotterdam: Balkema.
Potts, R. (1984). Home bases and early hominids. Am. Scient. 72, 338347.
Potts, R. (1988). Earb Hominid Activities at Olduvai. New York: Aldine de Gruyter.
Potts, R., Shipman, P. & Ingall, E. (1988). Taphonomy, paleoecology and hominids of Lainyamok. Kenya. J.
hum. Evol. 17, 597-614.
Shackleton, R. M. (1978). A geological map of the Olorgesailie area. In (W. W. Bishop, Ed.) Ceologicol Bnckcround
to Fossil Man. Edinburgh: Scottish Academic Press.
Shipman, P. (1986a). Scavenging or hunting in early hominids: Theoretical framework and tests. Am. Anthrop. 88,
27-43.
Shipman, P. (1986b). Studies of hominid-fauna1 interactions at Olduvai Gorge. J. hum. Bvol. 1.5, 691-706.
Shipman, P., Bosler, W., & Davis, K. L. (1981). Butchering of giant geladas at an Acheulian site. Curr. Anthrop.
22,257-268.
Toth, N. & Schick, K. (1986). The first million years: The archaeology of protohuman culture. Ados. Archaeol.
Meth. Theory 9, l-96.