Boxgrove: a Middle Pleistocene hominid site at Eartham Quarry, Boxgrove, West Sussex.

MCNABB, JOHN

M.B. ROBERTS & S.A. PARFITT (ed.). Boxgrove: a Middle Pleistocene hominid site at Eartham Quarry, Boxgrove, West Sussex (English Heritage Archaeological Report 17). xxiv+456 pages, 363 figures, 145 tables. 1999. London: English Heritage; 1-85074-670-2 paperback 80 [pounds sterling].

Boxgrove had a long history of Palaeolithic investigation, but the presence of in situ working floors was not suspected until Mark Roberts, the director, recognized the true significance of the site in the early 1980s. This volume present the details of the first phase of the project between 1983 and 1992. It precedes a much more detailed volume on the flintwork currently being written.

In general, the volume is very well presented and well written, with lavish illustrations and photographs of a very high standard. In conjunction with the recent British Museum monographs on Barnham and High Lodge, the Boxgrove volume sets the standard for publishing multidisciplinary Pleistocene research. It will be a tough act to follow -- but then very few sites get the level of funding Boxgrove enjoyed.

Chapter 1 sets Boxgrove into its regional and temporal setting. A critical part of this concerns the acceptance of the marine oxygen isotope record (OI). Chapter 2 details the geological and sedimentological history of the site with contributions on locating the buried chalk cliff, a long and very detailed section on the sedimentology of the stratigraphic profile, the beach gravels, mineralogy and the sediment micromorphology. The section is rounded off by a very useful summary for the general reader. Macphail's section is easily the most accessible and initiates an encouraging trend in the volume to include de-jargonized summaries of the significance of specialist researches on a unit-by-unit basis. This characterizes the best contributions in the volume, and it is to be hoped that this formula will be followed in subsequent volumes. The level of detail in these sections contributes to one of the most complete understandings of the depositional history of a locality during the Pleistocene yet known. In Britain, this has only been achieved in anything like the same detail at Barnham, Swanscombe and Hoxne.

Chapter 3 covers the palaeontology of the site with specialist contributions on forams and ostracods, molluscs, fish, amphibians and reptiles, and birds. Chapter 4 focuses on the mammalian record. One of the most significant features of the assemblage is its richness and the abundance of species represented. The care in recovery and processing of faunal samples has vindicated the time intensive approach employed. Preservation is strongly correlated with the extant cover of calcareous terrestrially derived deposits (Eartham Formation), which have preserved the fauna close to the cliff line. Both chapters 3 and 4 give a glimpse of the ecological complexity of the Boxgrove environment, particularly on the unit 4c palaeosol, where even localized micro-habitats can be identified from the bird and amphibian data, nicely supported by the sedimentological evidence. The quality of the science is reflected in the methodological maturity described by Stuart in his discussion on bird taphonomy, and Parfitt's working of the data into the wider European scene. Parfitt's mammal chapter is one of the most constructive in the volume, separating a detailed descriptive section from a unit-by-unit reconstruction of the environment aimed at the informed but non-specialist reader. It is a shame that the palaeontological section did not follow this lead, and that both chapters were not integrated into this kind of stratigraphic reconstruction.

Chapter 5 is dedicated to dating. This is one of the most critical chapters, and one of the few that left me with a certain disquiet. The problems revolve around an early (pre-Anglian) vs a late (post-Anglian) date. The majority of the radiometric age techniques, i.e. those dusted with the glitter of scientific objectivity, argue for a later date. The relative techniques, plagued by arguments centred on the validity of their comparative baselines, argue for an early date. The in situ horizons (4b and 4c) fall outside the reliability range for Uranium series and Luminescence. The sediments are not sufficiently magnetized for palaeomagnetic determinations. Electron spin resonance allied with U-series suggest an OI age within stage 7 (245,000 years). The amino acid racemization determinations place the site in OI stage 11 ([is less than] 423,000 years). Particularly disappointing for those favouring the early date is the suggested age range derived from marine coccoliths, also in OI stage 11. This is possibly an independent test on the dating, since the abundance ratios of these microscopic marine animals are ubiquitous. But no discussion is offered of lateral variability possibly present within these species, or of possible differences in abundance resulting from in-shore vs benthic situations.

Only one `scientific' method supports an early date, optically stimulated luminescence. Rhodes's complex natural normalized growth behaviours allow him to postulate an OI age in stage 13 (although, as he notes, the standard deviation could land it squarely in stage 11 or later). This section, which reads like a sorcerer's spell, makes it clear that the technique is still in its developmental stages. So the early date for the site rests squarely with faunal comparisons. A significant list of species present at the site extinct by the end of the Anglian is fielded. The most discussed indicator concerns the water vole Arvicola terrestris cantiana whose evolutionary trajectory is theorized to be complete by the later Cromerian. But how much is actually known about temporal and spatial variation in these small mammal faunas? On balance, the stage 13 date for the site appears the most reliable for the moment.

Chapter 6 is the archaeology. A brief overview of British Lower Palaeolithic archaeology contextualizes the archaeological discussions on quarry I (units 4b and 4c), quarry 2 (4c), the horse butchery site at GTP 17 and the so-called residual archaeology from contexts where the sediments with their archaeology have been moved from their original locations. Sections on the vertical movement of artefacts in Q2A unit 4c and the recognition of different percussors possibly used in tool making- at the site add considerable depth to this chapter. Laurels are clearly due for the sections on refitting (conducted by Bergman and by Austin). What is present is primary technological and typological interpretation, and this I found very plausible (with some exceptions -- see below). Primary data were presented, following a refreshingly simple descriptive methodology. It is a shame that the text (through no fault of the editors or contributors) is not able to offer the reader a clearer understanding of the stratigraphy and archaeology associated with the Eartham Form--mion than it does. Solid evidence for the occupation and character of environments in early glacial periods is sadly lacking in the British Palaeolithic record. It is to be hoped that a fuller discussion of the archaeology from these units will eventually appear elsewhere. The archaeology chapter concludes with a very useful review of the position of the Boxgrove hominid in terms of its European brothers and sisters, and a short discussion of the now famous tibia.

I only have two real criticism of this chapter. The first concerns layout. The excellent hominid sections appeared as if an afterthought, and would perhaps have worked better in the next volume. The archaeology is discussed on a locality-by-locality basis, moving from west to east, reflecting the importance of quarry 1 in the overall project. A unit-by-unit format would have made the archaeology easier to access. The current structure does not make it easy to get a feel for what was happening on the palaeo-surfaces. One other difficulty is the section explaining Newcomer's sub-divisions of biface manufacturing debitage. This is introduced in the Q1A section but is not properly explained until Q2A is discussed. The second problem concerns the level of primary analysis conducted. What is lacking is more detailed analysis exploring the relevance of the biface making data to the tool and raw material behaviour at the site. This kind of `behavioural technology in context' (i.e. primary level/small picture interpretation) would have been better included in this volume since it would have given a more rounded appreciation of hominids at the site. This lacuna does not detract from the quality of the end product, it just leaves it a little flat. I am sure this will be redressed in the next volume.

The current volume raises some interesting issues a propos the interpretations to be offered in the next.

1 It is clear that in the next volume a proactive hunting scenario will be offered to explain Boxgrove. This will involve complex co-operative behaviour, centred around `home bases' (p. 425) set away from the immediate vicinity of the site. Few archaeologists would deny some level of hunting and co-operative behaviour (the hominids are after all primates, and at Boxgrove they make tools to fashion other tools with -- soft hammers). But does the evidence at Boxgrove actually support this level of interpretation? It is based on the horse butchery in GTP 17 and other examples (e.g. a rhino pelvis). These show an abundance of cut marks, with animal gnawing and damage overlying and therefore later than the anthropogenic modification. The first access and carcass defence arguments are beyond dispute. But in themselves are these sufficient to prove the carcasses were acquired through proactive hunting constituting the main subsistence strategy? And is hunting the primary cause of the successful colonization of Europe? `[T]he decline of the large carnivores could be directly related to the colonisation of Europe by hominids, who reduced the available carcasses by being in direct competition with other carnivores' (p. 425). Anyone who has seen lions competing for carcasses in the wild would probably find this hard to swallow. The heritage of this early colonizer argues differently. Its African ancestry suggests a bipedal omnivorous generalist more able to cope with a variety of ecological possibilities. With intelligence and relational complexity selected for, this would be an animal able to hunt, but one that was equally at home as a scavenger. In this reconstruction hominids would compete with carnivores such as the small wolf found at Boxgrove, hyenas and the like. Carcasses could be defended against such competitors with rocks and pointed wooden staves. Large group size and co-operative action would be essential. The larger carnivores would be avoided. As for the nature of the spear wound in the horse scapula in GTP 17, I remain deeply sceptical. It has been noted that the nature of the hole is not inconsistent with a high-velocity puncture. It is necessary to model a thrown wooden spear, penetrating hide, muscle block, intervening tissue, and then punching a whacking great hole through the bone itself. Even for a big hominid such a feat would be very difficult, and the Boxgrove horse was very large. Why would it allow a hominid to get close unless it was sick? It would be mighty brave to tackle a rhino with only a wooden spear. I would like to see experimental data to support some of these contentions. To date, in the only such experiments I am aware of, a wooden point was driven through a horse scapula by a mechanical arm, a provocative but unrealistic test.

2 Activity differentiation is also alluded to and could be expanded on in the next volume. The only activity attested to in the Boxgrove record is carcass processing via biface manufacture, use and repair. Other activities must have been performed elsewhere. At High Lodge a core and flake industry with scrapers is stratified beneath bifaces, but the glacier that entrained them planed them off the same flood plain surface as the cores and flakes (Ashton et al 1992). Is this evidence of spatial patterning within subsistence/social organisation? At Barnham, biface manufacture in Area IV. 4 is located less than 50 m away from core and flake debitage on the same river bank in Area 1 (Ashton et al. 1998). At the Swanscombe Lower Loam knapping floor, a minimum of 9 nodules were brought to the locality already partly flaked. Flaking then occurred before the cores and some of the flakes were moved on (Conway et al. 1996). As with Boxgrove, this represents the movement of raw materials and artefacts embedded within isolated and episodic behaviour patterns.

3 The question of transport of artefacts and raw materials for knapping is particularly fascinating, and a strong argument in favour of complex organizational ability. Sceptics argue that the distances involved imply very little planning depth. But at the Swanscombe knapping floor there were 9 nodules being moved around the landscape. At Boxgrove, things are being made elsewhere and moved into excavated areas, and vice versa -- in one instance a stop is made to repair a biface. Even if this is only over a space of a few tens of metres, it is the scale of this movement, noted at every place where a large area was excavated, that is the real issue. I would urge the editors to develop this interesting line of argument.

4 Macphail's positing of the duration of the unit 4c palaeosurface encompasses the span of a single modern human lifetime. One modern lifetime may have been the equivalent of perhaps three or four Cromerian ones. If this be the case then the organization of behaviour was very stable from one generation to another: always the same thing, in the same way and in the same place. It is clear that the 4c palaeosol represents a dynamic environment for early humans and animals alike. The litter of debitage and abandoned bifaces lying on the ground would have been visible to hominids moving over that surface. Could such a constant visual trigger represent the stimulus for the monotonous regularity seen in biface form as well as explain the stability in behavioural practices?

5 We must be cautious about what we make of the Boxgrove data. As the editors are at pains to point out, there are no hearths nor any evidence of domestic activity, and just how representative of the organization of Acheulean lifeways the site is remains to be seen. We should be very careful not to set Boxgrove up as the flagship of Acheulean sites in Britain. What the site ultimately represents is a unique window on a tiny sliver of long ago, all the more poignant for lasting the short span of time between grandparent and grandchild.

References

ASHTON, N., J. COOK, S.G. LEWIS & J. ROSE (ed.). 1992. High Lodge. London: British Museum.

ASHTON, N., S.G. LEWIS & S. PARFITT (ed.). 1998. Excavations at the Lower Palaeolithic site at East Farm, Barnham, Suffolk 1989-1994. London: British Museum. Occasional paper 125.

CONWAY, B., J. MCNABB & N. ASHTON (ed.). 1996. Excavations at Barnfield Pit, Swanscombe, 1968-72. London: British Museum. Occasional paper 94.

JOHN MCNABB(*)

(*) Department of Archaeology, University of Southampton, Southampton SO17 1BI, England. scarab@soton.ac.uk