ERAUL 92, 2000, p. 139 à 145.


139


THE LAST NEANDERTHALS


Olga Soffer



Introduction

This volume, like it’s parent conference, focuses
on the “firsts” - the oldest this or that hominid
taxon somewhere in space. Although such a focus
has great antiquity in our field - it is also one
which addresses the least interesting and most
static question about the past: the “when”. This
chapter looks not at the “oldest” but at the
“youngest” or “last” archaic humans in Eurasia. I
do so because, as Mellars (l996) has pointed out,
questions about demographic histories and spatial
distributions of Pleistocene hominids have
received little attention in the literature. Yet, as
Howell (l994) notes, in order to understand human
evolution we must consider such factors as group
size, spatial and temporal continuity of occupation,
isolation, and extinction. These are basic biological
concepts that affect evolution through such
processes as the founders effect, gene flow, and
genetic drift. Without considering them we are
doomed to putting “dots on maps” and generating
static scenarios that offer few answers to the most
interesting questions about the past - why was
there change? why did evolution occur?

Change in geologic or archaeological time, be it
evolution or extinction, does not happen to
individuals - it happens to populations. All
populations, including Pleistocene hominid
populations, have histories: periods of success
when they expand and colonize new habitats as
well as periods of stress when they contract into
refugia. Sometimes refuging is temporary and
populations rebound. When no rebounding is
possible, local extinctions follow. Research on the
extinction of mammoths has shown that it is serial
local extinctions that ultimately bring about the
extinction of the taxon (Sher et al., l995; Soffer,
l993). The young dates for the last mammoths on
Wrangel Island, where they existed until some
4,000 years ago, reveal just how slow a sequence
of serial local population extinctions can be before
the demise of the last representatives of the taxon
occurs (Vartayan, l998; Vartnyan et al., l993).

The young Wrangel mammoths also inform us
about the relationship between refugia and
lifeways. Specifically, refugia are locations which
provide a species with a suitable niche - offering
stable environmental conditions for its way of life -
which permit populations to survive and compete
successfully.

Neanderthals and archaic lifeways

Recent research on prehistoric human populations,
from both the Pleistocene and the Holocene
documents that pre-Neolithic populations were not
stable in space or though time. Instead, from initial
colonization onward, all continents witnessed
settlement discontinuities including local
population extinctions (Lahr, l996, l997). Thus,
human history is comprised of multiple dispersals
and regional demographic instabilities.

The Pleistocene archaeological records from the
occupied Old World amply confirm this. During
the Last Glacial Maximum (LGM) some 20,000 -
18,000 years ago, for example, southwestern
France and the East European Plain served as
refugia for European populations (Housley et al.,
l997; Soffer and Gamble, l990). Likewise, both
Africa and Australia saw populations contract into
optimal habitats (Burzer, l991; Lahr, l997).

We see the same refuging phenomenon when we
look at the youngest Neanderthals and the last of
Middle Paleolithic lifeways. Specifically, I argue
that the adaptations developed by these archaic
hominids over many millennia gave them
competitive advantages in very specific regions of
Eurasia where they persisted until as late as some
27,000 years ago.

I underscore that in this discussion I neither
assume a 1:1 relationship between this taxon and a
particular technology. Nor do I assume that in the
last Neanderthals we see a taxon before extinction.
These are not Wrangel mammoths. These are
hominids with behavioral flexibility which could
allow for significant behavioral changes - for new
lifeways. That is exactly what we see happening
outside these refugia at such places as St. Cesaire
and Arcy (Harrold, l989; D’Errico et al., l998).
Thus, rather than discussing species extinction, this
chapter examines an end to a particular way of life.

1. Regional Distributions
Figure 1 plots the distribution of the last

O. Soffer


140


Figure 1. The distribution of the Neanderthal skeletal remains and Middle Paleolithic inventories dating < 35,000 BP in
Europe. Ì hominid remains; z archaeological remains.

chronometrically dated Neanderthal and Middle
Paleolithic sites in Europe. All date between some
35,000 and 27,000 years ago. The utilized
radiocarbon dates are uncalibrated - meaning that
with calibrations they would date from 2 - 4,000
years older in calendar years (Van Andel, l998
with references). Beginning in the west, they
include sites along the Portuguese littoral (e.g.
Gruta Nova, Figueira Brava), southern Spain,
Gibraltar, and the Spanish Levant (e.g. Gorham’s
Cave, Zafarrya) (Barton et al., in press; D’Errico et
al., l998; Straus, l997), possibly southern France
(e.g. Hortus) (Wolpoff, l996), central Italy (e.g.
Grotta Breuil) (Bietti, l997), southwestern Balkans
(e.g. Vindija) (Kozlowski, l996; Wolpoff, l996)
and Crimea (e.g. Buran-Kaya III, Kabazi II,
Starosel’e) (Marks and Chabai, l998; Pettit, in
press). Similarly young ages are reported from
Western Georgia (e.g. Akhshtyrskaia, Malaia
Voronovitsa) (Boriskovskij, l982, l989; Liubin,
l993), as well as from the eastern Eurasian sites in
the Altai and the Saian (e.g. Denisova and
Okladnikov Caves) (Derev’anko, l997; Kuzmin
and Orlova, l998).
The period in question falls in the second half of
Oxygen Isotope Stage (OIS) 3 - a stage from some
59 - 24 ky characterized by numerous very brief
sharp climatic oscillations including a well marked
warm episode at about 40,000 and a cold one at
30,000 years ago in the calendar year chronology
of ice and deep sea core data (Van Andel and
Tsedakis, l996, l998). Figure 2 depicts
reconstructions of biotic zones during an OIS 3
warm around 40,000 ky and a preceding cold
intervals. This cold interval, although preceding
the one at 30,000, can be used as its proxy. These
reconstructions shows that regions occupied by late
Middle Paleolithic populations were covered by
nemoral vegetation - a mix of broadleaf and
conifer arboreal growth which remained in these
areas, although somewhat reduced in extent,
throughout the cold stadials (Grichuk, l992). Van
Andel and Tzedakis’ (l998) reconstructions clearly
show that it is precisely these regions that served
as refuges for both deciduous and coniferous
species as well as for some Mediterranean
evergreens. It is important to underscore that
arboreal refugia existed here not only during this