. 32
( 43 .)


190 Neanderthals and Modern Humans

(7) If Upper Palaeolithic technologies evolved independently in Neanderthals
and were not the product of acculturation (d™Errico et al., 1998) then such
technologies should be found in Neanderthal regions of Iberia where there
was no contact with Modern Humans.

In this section I have shown how a local population effect may run counter
to a regional or more global population trend as part of the dynamics of the
system. Thus, within a geographical region opposing population trends may
result in different local populations and stochastic effects can be particularly
important in small populations. These observations raise the question of the
level of con¬dence which may be attached to the interpretation of patterns of
population spread, of transition or replacement of populations, from the study
of a few discrete and geographically widely-separated sites.
With this caveat in mind I now attempt to correlate the regional predictions
of the model for the ¬ve selected areas with the existing evidence from the ¬ve
representative sites. Since the dates used in the simulation are model-dependent
we can only compare site patterns relative to each other. For the end of the Mid-
dle Palaeolithic the model predicts the following sequence of disappearance:
Castillo, Ermitons/Carihuela; and Caldeirao/Gibraltar. The data in the literature
indicate the sequence Castillo, Ermitons, Gibraltar, Caldeirao/Carihuela (Vega-
Toscano et al., 1988; Vega-Toscano, 1990; Cabrera & de Quiros, 1996; Maroto
et al., 1996; d™Errico et al., 1998; Stringer et al., 2000). For the start of the
Upper Palaeolithic the model predicts the following sequence: Castillo, Ermi-
tons, Caldeirao, Carihuela, Gibraltar. The data in the literature indicate Castillo,
Ermitons (in fact sites in the same area), Caldeirao/Gibraltar/Carihuela (Vega-
Toscano et al., 1988; Vega-Toscano, 1990; Cabrera Valdes & de Quiros, 1996;
Maroto et al., 1996; d™Errico et al., 1998; Stringer et al., 2000). Given the
dif¬culties described above and the different dating techniques applied at the
different sites the observed trends would seem to match the model predictions
very closely.
The coastal region of Catalu˜ a would, according to the model, be a zone
of extended overlap. Maroto et al. (1996) record that ˜As far as Catalonia is
concerned, the passage from Middle to Upper Palaeolithic is characterised
by two phenomena which, though they may at ¬rst seem contradictory, are
quite compatible: abrupt replacement and coexistence.™ This is exactly what
the model predicts. In the Portuguese sites (Caldeirao and others) Raposo &
Cardoso™s (1998) data indicate some sites in which the Upper Palaeolithic and
Mousterian levels are in close approximation which could indicate proximity in
time and the Lagar Velho site (Duarte et al., 1999), as we have seen, is strongly
indicative. At Gorham™s Cave, Gibraltar, there is nothing so far that indicates
any temporal continuity between Middle and Upper Palaeolithic. This limited
Modern Human colonisation and Neanderthal extinction 191

evidence is suggestive but more work will be needed to test the prediction of a
variable mosaic of overlap zones, including regions where there was no contact
at all, within Iberia
Given an arrival into the open areas of Europe the Moderns would have found
little ecological or physical resistance to rapid spread and such progress would
only have been checked when reaching topographically or ecologically distinct
zones (in addition they probably did not meet many Neanderthals along the
way). Such areas would have included Alpine regions and other high moun-
tain blocks. The Iberian Peninsula, with its distinct biogeographical character,
would have provided such a check until climatic conditions permitted its rapid
penetration as the ecological barriers became increasingly diffuse. It is unlikely
that the spread of the Moderns would have come to a complete halt and the idea
of a barrier in the region of the Ebro (Zilhao, 1996; d™Errico et al., 1998) may
only be an illusion given the discontinuous distribution of archaeological sites
to its south and the dif¬culty of detecting initially small colonising populations
from such sites as exist. In fact, in areas of long-term regional sympatry pat-
terns such as those described for Iberian zones of overlap would be expected
with at least partly stochastically-driven local extinctions, replacements and
even apparent trend reversals occurring. Such variability has been described
for Middle Eastern sites (Stringer & Gamble, 1993; Bar-Yosef, 1994) and may
re¬‚ect a combination of the expansion of the southern limit of the Neanderthal
range in response to climatic deterioration and stochastic processes expected in
small metapopulations (Harrison, 1991; Foley, 1997; Hanski & Gilpin, 1997).
Stochastic effects would have also affected the Modern population which was at
low densities in North Africa and the Middle East during the early last glacial
period (Ambrose, 1998). In general, late Neanderthal populations would be
expected from heterogeneous environmental situations, at the regional scale,
where climatic characteristics maintained refuges of ˜non-steppe/tundra™ vege-
tation. Such areas, which at low latitudes would be expected to have the highest
mammal species richness (Kerr & Packer, 1997) and the greatest tree diversity
(McGlone, 1996), would have included, as we have seen, the three Mediter-
ranean peninsulas “ Iberia (Vega-Toscano et al., 1988; Antunes et al., 1989;
Zilhao, 1993, 1995, 1996; d™Errico et al., 1998; Raposo & Cardoso, 1998;
Finlayson & Giles Pacheco, 2000), Italy and the Balkans (Karavanic & Smith,
1998) “ the Middle East (Bar-Yosef, 1998), areas on the edge of the Russian
Plain (Soffer, 1989) and, to a lesser degree, the heterogeneous landscapes of
south-western France (Mellars, 1996). In this respect the more northerly dis-
tribution of steppe and tundra in eastern Europe and western Asia (Chapter 6)
than in western Europe (Mellars, 1996) could account for the late survival
of Neanderthals around the Black Sea. Such heterogeneous landscapes would
have reached their maximal extent at the start and end of interglacial conditions
192 Neanderthals and Modern Humans

(Mellars, 1996) and probably also locally during full interglacial conditions,
when forests spread across much of Europe, through the clearing action of
megaherbivores (Owen-Smith, 1988).
In Gibraltar, the emerging pattern is that of a Neanderthal population well
tuned into the seasonal resource cycles within their home ranges (Finlayson &
Giles Pacheco, 2000; Finlayson et al., 2000a). Neanderthals were omnivorous.
Opportunistic scavenging (Stiner & Kuhn, 1992; Mareau, 1998) would be an
additional tactic within such a system. Such a strategy would only be successful
in environmentally heterogeneous regions and would have reduced risks which
could have arisen from over-dependence on a single or small number of prey
types (Gamble, 1995). A similar situation applied in Italy (Stiner, 1994; Kuhn,
1995), south-western France (Mellars, 1996) and the Middle East (Stringer &
Gamble, 1993). The tactics of survival in regionally homogeneous landscapes
(such as the grass savannahs or the steppes) would have been very different,
requiring a different kind of social organisation, planning and long-range move-
ments (Chapter 5). A population with such a social structure (Bar-Yosef, 2000)
would have been predisposed to success. If we add to this the short duration
of the northern summers and the short daylengths of the northern winters we
see that a very different strategy would have been required to survive in the
environments of the Eurasian Plain during glacials. Such a strategy must have
included an improved monitoring of distant environments (Bar-Yosef, 2000).
The Moderns did eventually spread into the kinds of heterogeneous landscapes
occupied by the Neanderthals but they may have done so largely once the
Neanderthals had left these or after the optimal environments of the Moderns
had been ¬lled. The occupation of mountain environments by Moderns occurred
relatively late (Gamble, 1993).
Thus Neanderthals would appear to behave in the traditional sense as K-
selected whereas the expanding Moderns (with a growing population and
largely not at carrying capacity) would ¬t better an r-selected model (Chap-
ter 5; MacArthur & Wilson, 1967; MacArthur, 1984). McGlone (1996) has
suggested three ways in which temperate and boreal trees ran the risk of being
eliminated from large portions of their range during a glacial“interglacial cycle:
(1) through not colonising suf¬cient sites during their period of peak abundance;
(2) through elimination by stochastic effects or competition in long periods of
isolation; and (3) through failure to track suitable climates through migration
during periods of rapid change. The analogy for a K-strategy population of
Neanderthals is self-evident.
The main implication of all this then is that the arguments advanced in this
book that Neanderthal extinction and Modern colonisation can be understood
without recourse to interpretations that require the biological superiority of
one form over the other is reinforced. Quite simply, the Modern phenotype
was better suited to the variable and cooling conditions of OIS 3 and OIS 2
Modern Human colonisation and Neanderthal extinction 193

Europe than the Neanderthal phenotype was through responses to the expan-
sion of the biotopes with vegetation structural characteristics akin to those in
which that phenotype is likely to have evolved (Foley, 1987). The Neanderthals,
according to the model presented here, had been declining in Europe since
their heyday in OIS 5a, and especially during OIS 4, always in the absence of
The model establishes one point conclusively, whether or not Moderns and
Neanderthals interacted and whether such interactions were competitively un-
equal, such interactions are not a prerequisite for Neanderthal extinction. It is
perfectly plausible to explain the Neanderthal extinction as a result of envi-
ronmental change caused by climate as we have seen. As such there would be
areas of Iberia which would have been colonised by Moderns after Neanderthals
had disappeared. It is also possible to observe scenarios in which a declining
Neanderthal population and a growing Modern population could coexist and
inter-breed within the same geographical region for up to 4 kyr. These areas
would have been of reduced extent (1.5% of the Iberian surface area in this
model) so it is therefore unlikely that large-scale hybridisation would have been
the rule (Finlayson, 1999). A real measure of the degree of interaction between
the two forms would result from the study of independent cultural evolution vs
acculturation in regions where the two forms overlapped and others where they
did not at all.


I now summarise the events leading to the Modern colonisation of Eurasia and
the Neanderthal extinction.

Human adaptation

Humans during the course of the Pleistocene adapted to those features that were
relatively stable. These were the availability of medium-sized mammalian her-
bivores and the topography of the terrain. Neanderthals adapted to hunting
mammalian herbivores in broken terrain. Moderns adapted to hunting mam-
malian herbivores over ¬‚at terrain.


Increasing climatic instability fragmented the tropical African rainforests and
African hominids developed a biology consistent with the exploitation of patchy
194 Neanderthals and Modern Humans

environments. Behavioural plasticity and dispersal ability were favoured.
The Moderns were the ultimate expression. The Neanderthals were also be-
haviourally plastic, having descended from the common ancestor with the Mod-
erns. They coped with instability by extending the range of resources consumed
over a small area. Dispersal ability was reduced in comparison with the Mod-

Habitat tracking

Habitats moved with climate change. Moderns tracked intermediate and open
habitats and expanded from Africa. The expansion of open habitats in Eurasia
favoured this population. Neanderthals tracked intermediate habitats. As the
rate of change accelerated Neanderthals attempted to adapt behaviourally but
their morphology was a constraint in open landscapes. As their optimal habitats
contracted their range was fragmented. With low dispersal ability they became

The Modern Human super-organism

Moderns resolved instability through the uni¬ed action of teams. Social be-
haviour, language and symbolism bound teams together and enhanced informa-
tion ¬‚ow. Modern Human groups coverted unpredictability into predictability
by a variety of behavioural means. Behavioural adaptation, especially through
culture and technology, permitted them to ride the increasingly frequent envi-
ronmental oscillations. Moderns were also stressed during the LGM but they
managed to hold out and recolonised Eurasia and the world.
8 The survival of the weakest

The thrust of this book is that pressures and stresses on peripheral populations
in areas of population growth have driven the changes that have marked human
biological and cultural evolution. It has been the stressed populations that have
been the most innovative (e.g. Fitzhugh, 2001). Hominids have responded to the
increasing instability with risk reduction responses that ¬t under the umbrella
of increasing the spatio-temporal scale of operation. Correlates of increase in
scale have included, as we saw in Chapter 5, an increase in dispersal ability,
home range size, group size, neocortex size, the complexity of social behaviour,
symbolism, ef¬cient and mobile tool kits, and gracile morphology. Since global
climate and environments have become increasingly unstable over the last two
million years those marginal populations that adapted to local stresses in these
ways were able to turn disadvantage into advantage each time conditions dete-
riorated or became less stable. These adaptations would have evolved in periph-
eral populations that perceived marginal landscapes as spatially heterogeneous
and therefore spatially risky. These adaptations to exploiting patchy landscapes
then became advantageous in situations of increasing temporal heterogeneity,
that is in situations that were perceived as temporally risky. We can therefore
understand these adaptations as evolving through a normal process of natural
selection and we do not need to invoke alternative mechanisms, such as variabil-
ity selection (Potts, 1996a, b, 1998), to explain the observed patterns and trends.
Most of these adaptations would have been behavioural. However, directional
trends towards increasing patchiness also permitted morphological adaptations
such as limb morphology and neocortex size. Temporary, low-scale, climatic
and environmental improvements would have reversed the situation. We do
not know how many innovations subsequently became ˜extinct™ as no trace of
them would have been left in the archaeological record. The apparent halt of
the progress of the early Moderns in the Middle East after 100 kyr may be an
example of such a reversal. Had the global situation changed towards increasing
stability the picture today may well have been very different as conservative
behaviour would have been favoured over innovation. Let us brie¬‚y look at
large scale technological change to illustrate this point.

196 Neanderthals and Modern Humans

Technological innovation

Technology is the result of: (a) ecology, in response to the environment; and (b)
history, the product of traditions and their diffusion (Chapter 5). The changes
that we observe in the archaeological record, that take us from Mode 1 to
Mode 4 technology, do not re¬‚ect a linear evolution. Reversals, in cases of
the discarding of more sophisticated technology in favour of more rudimentary
forms are not infrequent. The adoption of a particular technology is the balance
between its costs and the bene¬ts gained from its use. A costly technology may
be adopted if it brings gains that far exceed the costs. If circumstances change
and the bene¬ts are reduced, the particular technology may be discarded in
favour of a less costly form. The apparent increase in technological complexity
through time is a re¬‚ection of cumulative historical effects and not of a linear,
progressive, technological evolution. What we do expect to observe through
time, in a world of increasing instability, is the development of technologies that
were increasingly costly to make but that were also increasingly bene¬cial in
unstable circumstances. What we would expect to increase is their ef¬ciency “
that is, the differential between the energetic costs of manufacture and the
bene¬ts derived from their use. I propose that the trend from Mode 1 to Mode 4
is one that goes from low cost/low bene¬t/low ef¬ciency technologies towards
high cost/high bene¬t/high ef¬ciency ones. We see this degree of sophistication
in the late Upper Palaeolithic technology, such as the light and highly specialised
microblades of north-east Asia and north-west North America that facilitated
the rapid re-colonisation of Siberia after the Last Glacial Maximum (LGM)
around 18 kyr (Goebel et al., 2000).


. 32
( 43 .)