The goals of the module concerned the characterisation of the Broom Lower Palaeolithic artefact assemblage, with specific reference to:

1. Technological and typological characteristics
The following discussion of technological and typological patterning in the Broom assemblage stresses intra- and inter-sample trends. Intra-sample patterning refers to trends within an individual sample, which could be the overall assemblage sample (n=640) or the below-datum sample (n=34 - see sections 2 and 3 below for explanation of these sub-samples). Inter-sample patterning refers to comparisons between individual samples. The following discussion is predominantly concerned with the biface element of the assemblage, given the high proportion of these artefacts and their greater potential for informative patterning.

The assemblage shows considerable intra-sample variation in all of the major categories (biface type, raw materials, blank form, tip and butt types, edge profiles, and size (weight). Although in many of these categories, there was a dominant type (e.g. cordate/ovate bifaces, medium-grained chert, irregular rounded tips, trimmed flat butts, and bifaces between 100-500g), the range of other types indicate considerable variation in technological practise and the end products of knapping activity. These patterns do not suggest strongly imposed standardisation with respect to the production of artefacts. This is also supported by the absence of inter-sample variation (discussed further below), evident in the stratigraphic sub-samples and the spatial sub-samples.

There is a strong preference for the production of straight edges, suggesting that the knappers’ were concerned with the production of good cutting edges on the bifaces. This is further supported by the dominance of trimmed or trimmed flat butts, suggesting the production of bifaces with fully circumferential cutting edges. The dominance of straight edges for all the major biface types at Broom further suggests that efficient cutting edges were required, regardless of the overall biface form (it is possible that the different combinations of cutting edges and biface forms might have afforded a wide range of task-related ergonomic tools).

In contrast to the claims made by C.E. Bean, asymmetrical (lop-sided) biface forms do not dominant the assemblage (Bean claimed that this type 4 represented c. 50% of his collection). No evidence has been found to support Bean’s claim that the asymmetrical bifaces were produced on side-struck flakes (and 12 of the asymmetrical bifaces were shown to have been produced on cobbles). Nonetheless, these forms are relatively common (19.5%, n=125), occurring on all raw material types. This implies that the asymmetry was not a consequence of raw material type. A wide range of biface types are represented on these asymmetrical forms, suggesting either that the asymmetry was an unintentional product of the knapping process (and that biface form was significant); or vice-versa (that the asymmetry was an intended end goal, and was considered to be more important than biface form). It is currently not possible to evaluate one interpretation over the other, although the dominance of straight edges on these forms suggests that they may have been no less functional than the ‘symmetrical’ bifaces.

The non-biface assemblage is dominated by retouched flake tools. Given the collection history at the site, it seems likely that this dominance of the retouched flake tools and unifaces over non-retouched débitage is a factor of their high degree of working and therefore unambiguous artefact status. The continued recovery of flakes in the 1980s (by C.P. Green) and the 2000s (by Hosfield & Chambers) also suggests that the débitage component in the Bean collection is highly under-represented. It is therefore not possible to reconstruct knapping behaviour in the Axe Valley. However, the presence of non-biface stone tools indicates that activities beyond those traditionally associated with biface use (e.g. butchering large animals) were being conducted within the demonstrated fluvial catchment area.

2. Temporal patterning
Division of the Broom biface assemblage into stratigraphic sub-samples was based primarily on the Bean archive and the previous work of Green (1988). Although it yielded relatively small samples (due to the number of accurately provenanced bifaces in the Bean archive), the samples suggested generally robust patterning. All three sub-samples were characterised by intra-sample variation that was similar to that of the overall assemblage sample (see point 1 above). However, there was very little inter-sample variation, suggesting little change through time in hominid behaviour (as represented by stone tools). In this respect, Broom is notably different to Swanscombe, with fundamentally different stone tool industries (Acheulean and Clactonian) lie in stratigraphic superposition within a single MIS-cycle (Conway et al. 1996; Wymer 1999). The OSL dates for the Broom sediments (see module 2) suggest that this time-span could represent as much as 20,000 years, raising the possible issue of each sub-sample representing a palimpsest. However, the compositional similarity of the three stratigraphic samples suggests that the intra-sample variability is not due to time-averaged over-printing, since it would require the same sequence of over-printing to occur three times, a scenario for which there are no analogues.

The modelled abrasion data for the three sub-samples shows a very similar pattern (the majority of the material appears to be derived from a local source catchment), suggesting that the spatial foci of hominid behaviour in the River Axe landscape (as represented by biface discard) remained consistent through time. The physical condition of the majority of the bifaces (e.g. edge damage) suggested that they were primarily transported and deposited in the coarse-grained gravels which occur throughout the Broom sedimentary sequence (including the Middle Beds), rather than being discarded in situ in the fine-grained sediments, as suggested by C.E. Bean.

As 82% of the assemblage could not be stratigraphically provenanced, it is possible that there is greater inter-sample variation at Broom which cannot be demonstrated. However, on the basis of the current analysis, there is very little evidence for inter-sample variation. It is also possible that the lack of inter-sample variation is due to the three samples being stratigraphically separated in the Broom sediments, but originating from a ‘single’, homogeneous occupation. We would argue however, that this hypothesis requires a complex model of episodic artefact supply into the fluvial system (either as a result of fluvial erosion of floodplain sediments or sediment movement from valley slopes). Further testing of this model will be undertaken in the next module.

3. Spatial patterning
Division of the Broom biface assemblage into spatial sub-samples was based on the artefact abrasion modelling work of Chambers (in prep). The samples suggested robust patterning, with a large concentration of bifaces derived from within a few hundred metres of the Broom pits. The remainder of the bifaces were derived from a wider source area, although there was very little material that could be demonstrated to have been transported from further than c. 1km upstream. These samples are described as on-site and off-site material respectively, although the terms are not strictly used in the Foley (1981) tradition. Both samples are characterised by intra-sample variation that is similar to that of the overall assemblage (see point 1 above). There was also (as with the stratigraphic samples) very little inter-sample variation, suggesting little spatial variation in hominid behaviour (as represented by stone tools). This would suggest an absence of a cultural geography (e.g. the existence of task-specific sites, the creation of taskscapes or the spatial differentiation of landscapes) (Binford 1980, 1987; Ingold 1993; Gamble 1999). In contrast, in so far as can be demonstrated by biface manufacturing techniques and discard patterns, there appears to be a high degree of behavioural homogeneity across the modelled landscape.

The apparent intra-sample variability in the on-site sample could be argued to result from over-printing and the generation of a palimpsest assemblage. However, we would argue that the absence of evidence for over-printing in the stratigraphic samples (and the overriding pattern of intra-sample variability) suggests that this is not the case. It is difficult to independently assess whether the on-site sample (and the off-site sample) represents a ‘single’ occupation or repeated visits to the landscape over a long period of time. However, the stratigraphic range of the artefacts in the Broom sediments suggests that the latter rather than the former is the case, supporting an over-arching model of behavioural homogeneity in time as well as space.

In conclusion, the analysis of the Broom archaeology has a series of archaeological applications:

1. That it is possible to extract meaningful archaeological information from derived, secondary context data sets, with respect to behavioural homogeneity/heterogeneity in space and time.

2. That assessments of the spatial and temporal origins of Palaeolithic artefacts in derived contexts can be made.

3. That it is possible to compare landscape archaeology associated with separate secondary context Palaeolithic assemblages.

Module 4 Outline

Module 4 Results

Module 4 Interm Report

References

Return to top