and transverse, with Quina retouch) dominate
the retouched tool assemblage, and resharpening
flakes indicate the onsite production and maintenance of these implements.
The elemental composition of 316 artifacts was
measured nondestructively using portable x-ray
florescence (pXRF) (31). The results indicate that
93.7% of the artifacts derive from the Gutanasar
volcano obsidian flows (2 to 8 km northeast),
2.8% from Hatis (12 km east-southeast), 3.2%
from Pokr Arteni (70 km west), and 0.3% from
Pokr Sevkar (120 km southeast) (figs. S15 and
S16). The latter two sources are located within
distant drainages not linked to the Hrazdan, and
therefore hominin transport is the only mecha-
nism to explain their presence at NG1. The pro-
curement of obsidian from a variety of local and
nonlocal sources suggests that hominins at NG1
were exploiting large, environmentally diverse
Early evidence for Levallois technology is found
in assemblages from Western Europe dated to
late OIS 9 and perhaps earlier (Fig. 1 and table
S6), but these are often from secondary contexts
or assigned to the “Final Acheulian” because of
the presence of bifaces and the low frequency or
absence of the preferential Levallois method. In
addition, these assemblages lack the Quina
scrapers that in part define the AY, where the
Levallois method is rare or absent (32). The lithic
assemblage from NG1 is unique in its combination of bifacial and Levallois technology, with
Quina retouch and blade production, all recovered from a secure stratigraphic context.
Given the absence of taphonomic mixing, the
intimate archaeological association of these technologies and artifact types could result from multiple hominin groups with distinct lithic traditions
occupying NG1 alternately over thousands of years,
thus producing a “mixed” lithic signature. However, this hypothesis would require us to accept
that LMP hominins were less technologically flexible than indicated by the African and Eurasian
archaeological evidence (6, 33, 34). Our data are
consistent with the hypothesis that the synchronic technological variance documented at NG1
reflects the behavioral variability and technological evolution of a local Late Acheulian population
and are thus inconsistent with the expectations
and assumptions of the single-origin and dispersal model for Levallois technology.
Empirical evidence supports the contention
that Levallois technology is an inherent property
of the Acheulian that evolves out of the existing,
but previously separate technological systems of
façonnage and débitage (7, 35), and shows that
Acheulian bifacial technology and Levallois technology are homologous, reflecting an ancestor-descendant relationship (36). Rather than a
“technical breakthrough” that spread from a
single point of origin, Levallois technology resulted
from the gradual synthesis of stone knapping
behaviors shared among hominins in Africa and
those indigenous to the Acheulian dispersal area
in Eurasia (Fig. 1). Consequently, the development of Levallois technology within Late Acheulian contexts represents instances of technological
The geographically and temporally discontiguous pattern of early Levallois technology and
the presence of Acheulian-like assemblages in
the LMP (≤late OIS 6) suggest that hominins
shifted between different technological options
and/or that technological change was not always
maintained, perhaps due to small effective population sizes, geographically restricted social networks, or high extinction rates (35). The eventual
proliferation of Levallois technology during OIS 8
to OIS 7 and its continued ubiquity into late OIS 3
(Fig. 1 and table S6) establish it as an evolutionarily significant adaptation practiced by diverse
hominin populations irrespective of taxonomic
affiliation or environment. As such, variations in
lithic technology cannot be considered proxies for
hominin demographic changes during the LMP.
At NG1, the early synchronic use of bifacial and
Levallois technology is consistent with the hypothesis that developments in the technological
realm of LMP hominins resulted from deep-rooted
evolutionary processes based on a common technological ancestry.
REFERENCES AND NOTES
1. I. McDougall, F. H. Brown, J. G. Fleagle, Nature 433, 733–736
2. J. J. Hublin, Proc. Natl. Acad. Sci. U.S.A. 106, 16022–16027
3. G. Clark, World Prehistory: A New Outline (Cambridge Univ.
Press, London, 1969).
4. N. Porat et al., J. Archaeol. Sci. 37, 269–283 (2010).
5. A. I. R. Herries, Int. J. Evol. Biol. 2011, 1–25 (2011).
6. C. A. Tryon, J. T. Faith, Curr. Anthropol. 54 (S8), S234–S254
1612 26 SEPTEMBER 2014 • VOL 345 ISSUE 6204
Fig. 4. Obsidian artifacts from NG1. Levallois: 1 and 2, recurrent cores; 3, 11, 13 to 15, and 17, flakes; 4,
point with retouched base; 5 to 8, blades; 9 and 10, preferential cores. Non-Levallois: 12, scraper with
Quina retouch; 16, biface.