Neanderthals and Modern Humans in the European Landscape during the Last Glaciation: Archaeological Results of the Stage 3 Project ed. TJERD VAN ANDEL and WILLIAM DAVIES
Cambridge: McDonald Institute Monograph 2003. ISBN: 1-902937-21-X. 265 pp (c. 110 figures) £35 hardback

The Cambridge Stage 3 Project has been one of the major research investigations of recent years. Stage 3, variously known as OIS 3 (Oxygen Isotope Stage 3) or MIS 3 (Marine Isotope Stage 3), covers the central period of the last glaciation from 59,000 to 24,000 years ago. This was the time when anatomically modern humans colonised Europe and when Neanderthals became extinct. This drama has been played out on a stage where the multiple changes of scene have been those of climate and environment. This book provides a degree of documentation rarely seen in the world of archaeology and asks pivotal questions about the nature of the changing environments and the interactions of humans, Neanderthals and fauna with the climate.

Structure
There are fourteen chapters which can usefully be grouped into overarching themes, viz

In addition to these themes there are small sub-plots and ‘cameo appearances’, as we shall see.

Epilogue
I very much wish that Tjerd van Andel, the project’s co-ordinator, had placed his excellent Epilogue at the beginning. His text provides a first rate and succinct overview of the results of the project as a whole and, so, would have focused the mind of the reader on the key trends and results. One must remember, in this context, that archaeologists are more likely to quarry this work for its particular themes than to read it from cover to cover. They are in danger therefore – to their great loss – of missing this overarching review.

Aims and objectives
Van Andel sets the scene in chapter 1 and explains that coverage was limited to the period of OIS 3 in Europe. Only existing data had been used and two key questions were identified at the outset (p. 1):
1. What was the climate of Europe like in OIS 3 and to what degree did the drastic changes displayed by the Greenland ice cores influence the European landscape and its flora and fauna?
2. Do the human events of the Middle and early Upper Palaeolithic reflect the OIS 3 climatic and environmental history and in what way and to what degree?

The work was carried out in two phases. The first of these involved collection of the environmental and chronological data. The second sought to map the changing presences of hominids and to use the data generated to examine issues of climatic tolerance and survival during the fluctuating climates of last glaciation Europe. Thus, Phase One was characterised as ‘climate and landscape’ and sought inter alia to test the models developed through such independent data sets as pollen and plant macro-fossils; loess, river gravels and permafrost features; and coleoptera. In all this, the detailed and well-documented Greenland ice-core record would serve as a key to interpreting climatic data, at least for the maritime areas of Europe. Phase Two of the project, entitled ‘the Palaeolithic in a climatic context’ sought to build up a ‘chrono-archaeological data base’ of all relevant dates between 60-20 cal ka BP. Note here that all 14C determinations have been calibrated to calendar years, using CalPal, so that they can be consistent with the calendrical ages provided by the Greenland ice-cores (pp. 5-6). It would have been helpful if some basic calibration tables, between 14C and calendar years, could have been provided for the benefit of the declining but significant numbers of readers who do not have access to, or cannot understand, computer-based calibration systems.

Chapter 3 provides a comprehensive discussion of the range of techniques that underpin the construction of chronologies for OIS 3, using a range of sources of which radiocarbon is but one. The possibility that radiocarbon determinations on bone, antler or horn may be ‘worthless’ is quietly introduced to worry us but is as quickly dropped, there being ‘no robust evidence’ (p. 28). This negative aside might have been better relegated to a footnote or omitted altogether.

Last glaciation environments in Europe
Chapters 2, 5 and 6 deal with last glaciation environments. Chapter 2 includes a valuable history of last glaciation climate in Europe. Here, the growth of an extensive Fennoscandian ice-sheet late in OIS 4 led to the formation of widespread tundra or cold steppe in Europe north of the Alps. The start of OIS 3 saw the initiation of a sequence of mild so-called Dansgaard-Oeschger events which were intercalated with cooler episodes. Long pollen and other environmental sequences at a number of sites, such as Les Echets or Grande Pile in France, or at Monticchio in Italy (fig. 2.4, p. 12), offer the possibility of regional terrestrial correlations with distant ice-core sites, combined with local information regarding plant and coleopteran successions from which inferred temperatures may be derived. The ‘early warm period’ (which becomes the ‘stable warm period’ in table 4.3, p. 33) ended about 45 ka. The climate worsened thereafter until, after a final warm event, a serious downturn began with temperatures broadly comparable with those of the Last Glacial Maximum (LGM), a period now seen as having been initiated c. 35 ka. One of the big questions asked in this book is whether the demise of the Neanderthals was in some way directly linked with this extreme climatic movement.

Two ‘gold standards’ of information relating to palaeo-climate in Europe are identified (p. 13). First, the Greenland ice-core GISP2 and, second, the terrestrial core from the Lago Grande di Monticchio in central Italy where sediments covering in excess of 100,000 years preserve a high resolution pollen record, well dated by a lamination-based chronology, with further support from tephrochronological and radiocarbon determinations. Chapters 5 and 6 develop the pattern of the environmental evidence for Stage 3 and show how the vegetational assemblages lack modern analogues and are not simply the result of latitudinal shifts (p. 92). I was intrigued (perhaps influenced by the Ents in Tolkien’s Lord of the Rings or by the movement of Abercromby’s long necked Beakers, changing their shape as they rolled north at the rate of two miles per year) to learn that, in the early Holocene, trees moved out of their once glacial environments at the rate of 0.2 to 2.0 kilometres per year (p. 93). The point here is that the trees may have moved at much the same speed during the warmer OIS 3 events.

The mammalian faunas of OIS 3
The chapters dealing with faunal issues (7, 10) highlight the many problems of identification, misidentification, and interpretation which await the researcher in this area. Rudolf Musil’s paper (Ch. 10) is focused on central and southeastern Europe. He makes the key point that many species are eurythermic and, so, are not strongly adapted to warm or cold climates. Indeed, the differences between species assemblages may relate more to contrasting maritime and continental climates. This author points up, too, the nature of both cave-assemblages (mostly carnivores and their prey) and open air accumulations (chiefly anthropogenic). There is a very useful catalogue of sites (pp. 171-81) followed by an analysis of diachronic change in faunal groupings. The inference is often made that mammoths were not hunted, except for situations like the ‘stampede’ at La Cotte on the island of Jersey. However, the fauna at Bohunice in the Middle Danube consists solely of mammoth remains and is interpreted as a specialist hunting site (p. 204). Of especial value to the archaeologist is Musil’s list of the principal faunal types in his area, giving Latin names, common names and ecology (table 10.2, pp. 189-90).

Neanderthal and modern human presences
Settlements and the use of caves
There is comparatively little material on the local geographical setting of sites, but this is more than balanced by the quality of the papers in this section. Here, I want to pay especial tribute to William Davies’s regional studies which bring a scale that is both local and human to this very European work. It was, perhaps, a bit surprising though to see the re-emergence of the ambush of game in small valleys concept (p. 211). I rather thought that Lewis Binford had put paid to the idea that large mammals are generally too stupid to anticipate that humans might be lurking in caves in narrow valleys (1978, 489-90). What Binford found, when studying the Nunamiut of inland North Alaska, was that it was pointless for the local people to ‘hole up’ in caves which, in that region tended to be located in side valleys, and simply wait for the game to walk by, because the animals knew perfectly well that the hunters were camped in the caves and simply kept away. Thus, it may be the absence of caves in the Lot valley (parallel to the valley of the Dordogne) relates less to factors of wind chill and temperature levels (Davies p. 206; van Andel p. 261) than to the unsuitability of such a relatively narrow valley as a place where hunting could be effectively pursued.

The distribution of settlements
Some valleys may have been inclement places but the thermal advantages of caves would have been well known to palaeolithic hunters and were probably not lightly neglected. The need for shelter – rather than ambush – may explain, therefore, the concentration of occupation of side valley caves in the Ardennes, the most northerly of Davies’s study areas. Rather more needs to be known, however, about the micro- and macro-climatic contexts of caves, their terroir, and of the wider terrain in which they were situated, in addition to the evidence of the use of these caves by humans and other carnivores. It is clear that there were factors which led to settlement becoming strongly focused in one region, whilst contiguous sites and areas were less used or not exploited at all. One suspects that in some cases a sort of law of ‘primogeniture’ applied: thus, last glaciation populations were globally small and seem to have been growing from perhaps a few tens of thousands at the start of OIS 4 to around a million in the early Holocene (Lahr and Foley this volume, p. 241; Mithen 2003, 11). Thus the population of a particular area – whether occupied seasonally or not – might have grown intra regionem without societal fission leading to the colonization of new areas. The human fear of the unknown is very familiar to us all and the quotation from Romeo and Juliet that there is ‘no life without Verona’s walls’ speaks to us as surely now as it did to audiences of Shakespeare’s day. Pari passu, there may have been some wholly disjunctive movements in which groups that had broken away leap-frogged adjacent areas empty of settlement, before reaching their ‘promised land’. Indeed, something of this kind has been argued for the earlier Aurignacian in Europe (Davies 2001).

Human presence in OIS 3
Chapter 4 looks at the distribution of Neanderthals and anatomically modern humans within a climatic range that was mild at its beginning c. 60,000 years ago, but was in a state of serious deterioration as the Last Glacial Maximum was approached. I would suggest that the authors’ use of archaeological industries as proxy evidence for Neanderthals or anatomically modern humans, whilst well and honestly handled, is not without risk. Here, it is worth noting that Aurignacian artefacts have twice been found in association with Neanderthal remains, at the sites of Trou de l’Abîme in France and Vindija in Croatia (p. 31 and table 4.1, p. 32). There are no instances, however, of the reverse phenomenon. The publication – since the appearance of this book – of new radiocarbon determinations from Vogelherd, SW Germany, has further undermined faith in the interpretation of the Aurignacian as a product of anatomically modern humans rather than Neanderthals. The background to this story is that Vogelherd had produced – apparently from well-stratified contexts in layers V and IV – a dozen examples of figurative art, rich Aurignacian assemblages, and a last glaciation fauna (Conard et al. 2004). These layers were well dated to between 36-30,000 14C BP but new dates on four human specimens have yielded wholly unexpected late Neolithic results of c. 5-4,000 14C BP, and show the human remains to have been intrusive. Indeed, Cro-Magnon itself has now been dated to 27,760 14C BP and the apparently well-stratified site of Mladec in the Czech republic lacks dates on the human remains. Modern human remains have been dated to ~35,000 BP at Pestera cu Oase in Romania but lack archaeological associations. There are, therefore, now no certain associations of Aurignacian assemblages and Modern Human remains in Europe and, thus, Conard and colleagues argue that it is equally plausible that Aurignacian industries were made either by Neanderthals or by Modern Humans

Neanderthals and Modern Humans: climate and range
The scenario of settlement was played out against two major climatic trends: N/S, from arctic to cold on the northern edge of the east-west European mountain chain, and W/E from the maritime Atlantic to the plains of continental Europe. Time-based trends show the picture set out below (based on table 4.3, p. 33). It might have been useful had the names given to the climatic phases been more easily conveyed to memory.

The progressive climatic downturn from 37 ka marked the onset of Neanderthal decline, just as the warm upturn from 59 ka saw a wide Neanderthal spread south of the 50°N parallel. Indeed, we find repeatedly in this book a consensus that Neanderthals were distinctly unenthusiastic about the cold. The Neanderthal decline involved a double seawards withdrawal, west to the Atlantic and southeastward to the Black Sea, the latter perhaps triggered both by climatic change and by decline in the availability of game on the Russian plains (p. 39). The issue of whether Aurignacian = Modern Human remains plagued with doubt but accepting this equivalence offers the following picture. First, that the patterns of movement of both Neanderthals and Modern Humans during the period 45-37 ka reflect similar climatic preferences. From this the inference might be drawn that both groups were adapted to the hunting of largely sedentary game and, so, neither population might have easily been able to adapt to the high level of seasonal movement seen recently under truly arctic conditions. A surprise of a different kind is the suggestion of the ‘dual nature’ of the Aurignacian with colonisation routes into Europe via both the Danube and Gibraltar (p. 44). This is clearly problematic but the idea is challenging and, as the text says, ‘worth evaluating’. ‘Early Aurignacian’ expansion had stabilised by 35 ka and, so, at roughly the same time (38-35 ka) as the appearance of the very earliest Gravettian: this appeared, in central Europe, at the sites of Dolní Vestonice and Höhle Fells; in Bulgaria, at Temnata; and, in Russia, at the sites of Mezmaiskaya and Kostenki. The authors of chapter 4 (van Andel, Davies and Weninger) hazard a guess that the origins of the Gravettian may have lain in Moravia, where it may have developed from the Aurignacian. What is strikingly different where the Gravettian is concerned is that the people who made it display demographic patterns which stand apart from both the Mousterian and the Aurignacian. The makers of the last two assemblages named seem to share many behavioural patterns (and so, perhaps, belonged to the same Neanderthal species?) but the makers of the Gravettian did not decline as the climate fell away; rather, they increased, notwithstanding the most adverse of circumstances, and eventually reached a plateau at 25 ka. Moreover, whilst not adverse to residence in the milder climes of Italy, they seem to display a marked preference for the colder north, up to 67°N, and the central European Alpine foreground, with distributions that were at times far denser. Indeed, in Moravia, the sites of Dolní Vestonice, Pavlov and Milovice, have often been regarded as large permanent settlements. The chapter closes with a series of questions largely concerned with Neanderthals: particularly embracing issues of origins, migration routes, relationships between dispersal patterns and climate, as also the question of responses by Modern Humans and Neanderthals to the same climatic situations. Final questions relate to Neanderthal extinction, the advanced adaptation reflected in the distribution of Gravettian sites, and the issue of settlement placing and spacing (pp. 49-50).

‘Snow, snow, thick thick snow’
The words that act my title for this section were useful (just) to people of my own generation when learning the rhythm and steps of the waltz. For the early peoples who inhabited Europe, snow was at times a matter of life and death. If it masked landscapes, food might be denied either to the game animals on which the humans depended or, directly, to the humans themselves. It might be necessary for humans to become not ‘man the hunter’ but ‘man the shoveller’ (Gamble 1987), in order to locate or recover carcasses buried in the snow, rather like a squirrel looking for his nuts. An important part of this book (notably chapter 8) is concerned, therefore, with a series of climatic indicators among which the issue of the thickness and density of snow cover, and its impact on human and animal life, looms large.

The first half of OIS 3, from 59-37 ka, was relatively warm. The downturn was initiated in 37 ka; by 30 ka it was seriously cold; and 10,000 years later, the Last Glacial Maximum was at its height. The variables most relevant to the condition humaine are summer and winter temperatures, day/night contrasts, wind-chill, and precipitation (which includes rain, snow, hail, fog and mist); important also, as far as snow is concerned, are both thickness of snow cover and the number of annual snow days. The focus here is on both the tolerance and preference of humans for a range of climatic conditions. These are assessed separately for the three archaeological techno-complexes which spanned the period 60-20 ka, namely Mousterian, Aurignacian and Gravettian. There are problems, however, in doing this for we often cannot be sure whether the location of a site, at a particular latitude or in a particular place, reflects year-round or seasonal human presence. The word ‘aestival’ creeps in here and seems to be used as an adjective pertaining to summer. But its potential use is wider for its primary meaning is heat (Latin aestus) from which the summer may be inferred. It also has the sense, when used zoologically, of heat-induced torpor. The latter sense would however convey the opposite of that intended, for those aestivating are our vigorous cold-loving Gravettians. Closing this linguistic diversion, we may note – here particularly commending the editors/authors for the regular points of note sections that make navigation through these frequently ice-laden waters so much easier – that the wind chill and snow cover tolerances of the makers of all three techno-complexes are very similar. Tolerances of mean temperatures present, however, a slightly different story. Mousterian and Aurignacian tolerances of temperatures and wind-child factors are actually very similar but, with the Gravettians we find, for the first time, that a distributional shift has occurred wherein most sites are in regions where winter temperatures are well below 0°C. The difficulty here, of course, is that we are not really sure whether we are looking at sites where the occupation was limited to the summer (what one might call ‘aestival’). But we need to look in more detail at the issue of snow.

My paternal grandmother, an otherwise wonderful woman, believed that leaves fell off trees purely for her discomfiture. Yet I look back to the same period my memory is quite other, and the image that remains is of a time when I enjoyed some spellbinding autumnal walks, crunching underfoot deep carpets of red and brown deciduous leaves. Let us not, therefore, see snow depth purely in negative terms as a constraint ‘through its impact upon prey, inhibiting large mammal grazing and mobility if it became excessive’ (p. 141). Snow provides readily accessible supplies of drinking water. Snow can also be fun. We can see this from our primate context where macaques of the Joshinetsu Plateau in Japan are known to have fun playing with snowballs (Ford 2000, 5). In Europe, the evidence for not just palaeo-spelaeology, but palaeolithic palaeo-spelaeology, is powerful not just in Magdalenian contexts but, even earlier, before 30,000 years ago (Clottes 2003). People may have had several reasons for venturing into the northern wastes: perhaps economic (food or ivory); ritual visitation as a socio-symbolic act; or exploration. These movements may have been prompted, however, by the stress engendered through being part of annual social aggregations. I have suggested elsewhere that small scale movements of hunter-gatherers may sometimes have involved ‘holiday groups rather than hunting groups’ seeking to chill out away from areas of social stress (Aldhouse-Green 2000, 41). Elsewhere in the volume here reviewed, the authors seem to suggest that occupation at several cave-sites was suspect because of the contingency of alleged human presence with extreme climatic conditions. Thus, doubt has been cast on the validity of the Aurignacian presence at Uphill Cave 8, inferred from ‘a solitary dated bone point’ (p. 143) and the fact that the Mousterian site at the Hyaena Den, Somerset, enjoyed the coldest wind chill temperatures in Europe (p. 155) has been identified. We have to remember several things here. First, we do not know the season of occupation at either site and that even the summer temperatures of the Last Glacial Maximum were mild (p. 259). Second, both sites are caves, precisely where you would expect to find visitors to the ‘frozen north’ sheltering from the lowered temperatures and wind-chill of that northern world. This is not the place to explore these ideas further, but I will be developing these issues in publications elsewhere. The conclusions advanced in chapter 8 are broadly as follows:

• Neanderthal snow preferences are consistent with less than 5 centimetres of snow depth and less than 60 days of snow cover per year
• Neanderthal tolerances reach no more than 50 centimetres of snow depth and 210 days annually of snow cover
• only as Neanderthals approached extinction do tolerances increase to 100 centimetres of snow depth and 240 days of snowcover (‘Hobson’s choice’?)
• the Aurignacian pattern is similar to that of the Neanderthals
• from 37 ka, both the Aurignacian and Gravettian sites show preferences for snow-depths of up to 20 centimetres
• during the Last Glacial Maximum, Gravettian depth tolerances reached 150 centimetres

These extremely valuable analyses on the issues of Neanderthal and modern human presences set the scene for a wider discussion of Neanderthal thermoregulation and extinction. Overall, the data seem to show is a high degree of similarity between the early Aurignacian and the Mousterian, such that ‘the earliest Aurignacian has more in common with the contemporary Neanderthals than with the later manifestations of the same techno-complex or with the Gravettian’ (p. 145).

Neanderthal thermoregulation, climatic stress and extinction
The first of three chapters on the above theme is chapter 9, written by Leslie Aiello and Peter Wheeler. They make their starting point a recapitulation of three ‘rules’ relating to the relationship between body form and climate (p. 147) viz:

• Bergmann’s rule species living in colder climates will have a higher body mass than populations elsewhere
• Allen’s rule species living in colder climates will have shorter extremities than populations elsewhere
• Ruff’s rule species living in colder climates will have very broad trunks when compared with populations elsewhere

Neanderthals correspond to the above with a very wide bi-iliac breadth (like the modern Inuit), relatively short legs and distal extremities expressed in the following ratios: lengths of radius:humerus, tibia:femur, and femoral length relative to femoral head size. Also suggested as signs of cold adaptation are large noses, large paranasal sinuses and large brains. The data studied by Aiello and Wheeler relate specifically to body form. The key result is that there is only a comparatively small difference between Neanderthals and Modern Humans as far as adaptation to cold is concerned. The Neanderthal advantage amounted to no more than 1°C before thermoregulatory thermogenesis would have to be initiated; a difference of 2.5-3.5°C applied in the case of their minimum sustainable ambient temperature. The hypothesis that Neanderthals might have benefited from a thick layer of insulating body fat is considered and rejected, on the basis that the weight of fat needed to produce a meaningful result would be more than half of an individual’s calculated body weight. The possibility of hairiness could not be evaluated but it is clear that Neanderthals would have needed the provision of shelter, fire and some kind of clothing. What is perhaps surprising at first sight is that median winter wind chill temperatures calculated for individual archaeological sites vary from a ‘warm’ +3.1°F (Mousterian), through -4.1°F (Aurignacian) to a distinctly chilly -8.7°F in the case of the Gravettian. But this may all be a reflection of an impaired Neanderthal capacity both to design and make clothing and perhaps also in maintaining and enhancing a high energy diet.

Human dispersal, migration and evolution
Two key events happened in Europe in OIS 3: the extinction of the Neanderthals and the arrival of Modern Humans. These are considered by Marta Lahr and Robert Foley (chapter 14). Their stimulating discussion of evolutionary processes cannot be covered here. However, it is clear that anatomically modern humans did not expand into Europe before 60 ka. Their analysis of Stage 3 events in Europe has some significant conclusions:

• there is no relationship between climate and the number of archaeological sites as a whole
• but, if Middle Palaeolithic sites alone are considered, there is evidence that Neanderthals were adversely affected by climate
• a refugium zone existed at 43-45° of Latitude (the level of southwestern France) but the more intensive use of this in Stage 4 rather than Stage 2 suggests that Neanderthals had a lesser tolerance of cold than Modern Humans
• that genetic data, whilst immensely valuable, is less fine-grained than archaeological and fossil evidence; accordingly, genetic data seem unlikely to be of significant use for resolving key questions regarding the interaction of human demography and climate

Presentation
Overall presentation is of a remarkably high standard. Even so, I find it remarkable that there is no index. Perhaps this reflects an ancestral memory of early BARs. Even Mortimer Wheeler – who ‘inter arma [had] no heart for studentship’ – managed in wartime Britain to prepare an index for his monumental work on Maiden Castle. It is a pity, too, that there is no composite bibliography which would have been considerably more useful for the reader than simply locating references at the end of each chapter. The effect of these two lacunae is to make interrogation of the data more difficult to achieve. I am reminded of the famous quotation from Nennius’s Historia Brittonum ‘Ego … coacervavi omne quod inveni’, which may be loosely rendered ‘I have made a heap of all I know’. With so many figures, problems were likely to arise and here we may note, particularly, fig. 9.5 on p. 156, where symbols have not reproduced clearly, and fig. 6.12 on p. 98, where the incomplete key renders interpretation of this map difficult. A very few other such points include Appendix 10.1 on pp.186-89, which began its life as Appendix 10A; a curious statement on p. 221, suggesting that Neanderthals and Modern Humans co-existed in Cantabria until ‘10 ka BP’ (I suspect that an editor has slipped in an unwanted ‘BP’); finally, the dates of 60-50 Kyr in the top image of Fig. 14.3 purport to relate to OIS 4 human dispersals but it might have been better to have said ‘after Stage 4’.

Reviewer’s conclusion
The book overall is a triumph of investigation into a very important part of the last ice age. Perhaps predictably, the project proved so successful on such a range of fronts that ‘we abandoned the idea of a complete and systematic report and left the authors free to select their topics’ (van Andel, p. 257). In answer, to Tjerd van Andel’s final question (p. 262), this work is clearly overture rather than finale. Congratulations to Tjerd van Andel and William Davies, and to all the contributors.

Stephen Aldhouse-Green
University of Wales,
Newport

References
Aldhouse-Green, S.H.R., 2000. Palaeolithic and Mesolithic Wale, in Lynch, F., Aldhouse-Green, S.H.R. and Davies, J.L., Prehistoric Wales. Stroud: Alan Sutton, 1-41
Binford, L.R., 1978. Nunamiut Ethnoarchaeology. London: Academic Press
Clottes, J. (ed.), 2003. Return to Chauvet Cave. Excavating the Birthplace of Art: the First Full Report. London: Thames & Hudson
Conard, N.J., Grootes, P.M. and Smith, F.H., 2004. Unexpectedly late dates for human remains from Vogelherd. Nature 430, 198-201
Davies, W., 2001. A very model of a modern human industry: new perspectives on the origin and spread of the Aurignacian in Europe. Proceedings of the Prehistoric Society 67, 195-217
Ford, B.J., 2000. Sensitive Souls. London: Warner Books
Gamble, C., 1987. Man the shoveler: alternative models for Middle Pleistocene colonisation and occupation in northern latitudes, in Soffer, O., (ed.), 1987. The Pleistocene Old World. New York: Plenum, 81-98
Mithen, S., 2003. After the Ice. A global human history: 20,000 to 5,000 BC. London: Weidenfeld & Nicolson

Review Submitted: August 2004

The views expressed in this review are not necessarily those of the Society or the Reviews Editor.