Figure 34: Measured flake scattergram; for key refer to text.
The lithic assemblage presented for study was derived from the 'plough-zone', 'topsoil' and 'excavated features' and was collected in three seasons of excavation, 1972, 1973, and 1977. There is no direct record of distribution across the site other than any zonal patterning that might be derived from the year of excavation and trench description. In total, 179 pieces of worked material were individually described in a single inventory (see archive); a single piece was identified as a chert and the remainder was flint. The assemblage composition is summarised in two parts: that derived from excavated features, and that from the plough-zone and topsoil (Table 3).
This was a small assemblage, with roughly equal numbers recovered from the plough-zone (83) and excavated features (96). The differences in composition between the plough-zone and excavated features were of interest, in that the number of broken flakes was not greatly different, suggesting that ploughing was not the general cause of breakage. The number of spalls collected from excavated features indicated some differential collection, as also did the small proportion (by number) of cores. However, the essential similarity in proportion between these two units of analysis suggests that the assemblage could be treated as a single unit.
The analysis of the assemblage immediately identified a strong Mesolithic component, in the presence of blades and microliths. However, it was also apparent from the flakes and tools that later industries were also present, and elements of these were stratified in association with Neolithic and Bronze Age pottery. There was no indication that there were any later elements and the material can be understood as residual when in later Iron Age features. In the analysis of the assemblage the wide variation in the body colour of the worked flint was recorded and compared with the Mesolithic assemblage from Lightmarsh Farm, Kidderminster, c. 2 miles (3.5km) north of Blackstone (Bevan 1994) and on the same side of the river.
There was a single piece of chert in the assemblage, and the remainder was flint with a considerable variation in body colour. A thin-rolled cortex, present on 42 pieces, showed little variation and was not studied further. The source of the flint was local; the secondary geological deposits and gravel terraces of the Severn river valley. The colour (Tables 3 and 4) varied from a beige (B) to a cream (CR), a light grey (LG) to a medium (MG) and blue-black (BB), a white-patinated (WP) and yellow-patinated (Y) surface or body. The only body colour not present in the Lightmarsh Farm Mesolithic assemblage was the blue-black component. Few beige and yellow elements were present in the Blackstone assemblage.
Flakes were the largest component of the assemblage (150), of which more than half were broken (80). Of the broken pieces a considerable number were probably derived from small blades (31). Of the complete flakes there were 33 tertiary, 32 secondary trimming and 5 primary flakes. Secondary retouch was only present on eight tertiary flakes and five of these exhibited side retouch, of which one was serrated; of the two with retouched butts, one also exhibited side retouch and one was notched.
There were sufficient complete flakes (70) to measure for a statistical analysis. The scattergram of breadth:length (x:y; Fig. 34) was plotted for each material colour, described above, to test the possibility that the Mesolithic, Neolithic and Bronze Age components might be so differentiated in the assemblage. Barfield (2007, 106) has suggested that in the West Midlands the colour and other macroscopic criteria of worked flint assemblages might suggest differential flint sourcing through time.
Figure 34: Measured flake scattergram; for key refer to text.
The scattergram shows a cluster of small blades and long flakes (6) within a length range of 10-30mm and a breadth range of 5-12mm, of which 50% have a white-patinated (WP) body colour (Fig. 34, Zone A). It is the distribution of the WP flakes (16), 33% of which are clustered in the small blade and long flake zone, that most closely matches the scattergram of the Lightmarsh Farm and other Mesolithic assemblages (Barfield 2007, fig. 7). Of the broken flakes, 31 were probably fragments of small broken blades and of these 17 were WP broken blades. This WP material was perhaps preferentially selected or was more available in the Mesolithic period, although it is clear that other flint was also utilised. The patination is post-depositional.
Of the other material-flake distributions a number of observations could perhaps be of interest. There were no blades (length:breadth of greater than 3:1) of light-grey (LG) flakes, although there were a number of small LG flakes within the 'Mesolithic cluster' of WP flakes. This LG material formed the largest component (50%) of squat flakes (length:breadth of less than 1:1). The distribution of LG material away from long flake and blade production perhaps indicated the LG flint was a larger source component in the later Neolithic and Bronze Age. Similarly it might be argued that the blue-black (BB) component had a stronger association with the Neolithic, with fewer broader flakes in the scattergram distribution (Table 4 and Fig. 34).
Table 4: Measured flakes, proportion of WP, BB and LG body colour (source material; see text for explanation)
| white-patinated | blue-black | light-grey | All | |
|---|---|---|---|---|
| blades | 4 | 2 | 0 | 8 |
| long flakes | 3 | 5 | 3 | 18 |
| totals | 7 | 7 | 3 | 26 |
| 30% | 30% | 10% | 100% |
Although there appeared to be a number of trends in this analysis that may indicate there were a number of industries present within the Blackstone assemblage that had preferred sources and a preferential selection of material, all the material was derived from the same source, the secondary geological deposits and gravel terraces of the Severn river valley. The statistical numbers here were also small, and so more exhaustive comparisons would need to be undertaken with assemblages from other sites, beyond the scope of this study.
Figure 35: Flint.
Late Mesolithic activity at Blackstone is clearly identified in the small blades and microliths in the lithic assemblage. All the material used was local to the Severn valley. The material is likely to have been won from river bank erosion and tree-throw exposure, with the quarrying of the root hollows into the river terrace deposits. Although the scoops on the site (Fig. 5) might represent the last traces of such an activity, the presence of only one small-blade core in the assemblage does not suggest that there was blade production on the site; rather the blades had been carefully curated to the site for tool production, as was also indicated by the large number of small broken flakes and spalls. Such a site in the river valley, rich in game and close to a river flowing with fish, may well have been a temporary hunting camp, possibly with a constructed shelter with a dry floor in the scooped hollow of a tree-throw, a natural wind-break and shelter, with access to a plentiful supply of wood.
Early Neolithic activity on the site can perhaps be identified in the longer and larger blades, core rejuvenation flakes and the rather broken core material; compare the opposed platform core that came from a large piece of raw material. With woodland clearance and an increased pace of ground disturbance for agriculture and grazing, the exposure of raw material in larger pieces would have been widespread, and the choice more varied. Preferred blue-black flint may have become more available in larger pieces, with the production of larger tools. The broken fabricator and small scrapers can be associated with this Mesolithic-Neolithic transition, as can the knapped flakes in pit 1027 with its associated Neolithic pottery.
Later Neolithic and Bronze Age activity on the site could perhaps be recognised in the presence of a wider and broader flake component to the assemblage that broadens the scattergram of all measurable flakes. Fine invasive retouch, present on one scraper, is a Beaker period technique (Fig. 35, no. 5). The latter and a thumbnail scraper (Fig. 35, no. 4), could be assigned to specialist tool production of the early Bronze Age period. There is also a sense of random production in the large variation in the platform depth (up to 7mm) and thickness of flakes (one was 18mm in thickness) from unprepared cores, suggesting a less specialised tool production of the middle Bronze Age. This latter aspect has not been exhaustively analysed, but with so few tools and datable features a detailed study would perhaps be unproductive. The butt-trimmed larger scrapers might have been associated with this later industry.
The excavated assemblage was derived from a range of features, a number of which can be assigned to the pre-Iron Age: 0015, 0016, 0023, 0079, 0083, 0113-17, 0120, 0149, 0183, 0552 and 1512-15 (see Site Description, and Fig. 5). The remainder were from those structures assigned to the Iron Age, principally the enclosure ditch.
Analysis (Table 5) shows successive residuality within later settlement features. However, 60% of the assemblage could be assigned to the pre-Iron Age, and of the remainder only 21 flints (23%) of the stratified assemblage were not assigned and were residual in the Iron Age, and of that material three were spalls. The component from Iron Age features is also equal to and of the same size, proportion and character as that from features assigned to the pre-Iron Age. Thus it is clear that the component in Iron Age features was residual and there is a measurable 23% residuality on the site.
Table 5: Site distribution of flint by period of feature.
| pre-Iron Age features | Iron Age features | unstratified | % | |
|---|---|---|---|---|
| broken blades, WP, Mesolithic | 19 | 10 | - | 32 |
| blades and long flakes, Mesolithic/Neolithic | 10 | 10 | - | 22 |
| other Mesolithic/Neolithic/early Bronze Age | 2 | 2 | - | 5 |
| spalls and other flake material | 17 | 21 | - | 41 |
| total | 48 | 43 | 4 | 100 |
The single piece of chert, perhaps from a source further afield or an erratic from wider disturbance across the river valley, might be derived from any of these industries.
Although the detailed analysis of the measured flakes does indicate that there was a differential sourcing of lithic material through time in prehistory, the assemblage studied was small. However, the potential for study and comparison with larger stratified assemblages in the Severn valley is clear, but winning that information will be difficult without stratified assemblages with clearly defined residuality; undoubtedly the winning of detail on lithic-sourcing will be at least as difficult as it was sourcing particular material for tool production in the Mesolithic. Hopefully, the task will clarify as more and better collections become available for study, just as flint sources became more available in the Neolithic and Bronze Age periods. The analysis was also a demonstration of the fact that surface collection from the plough-zone will define small concentrations of settlement activity on such sites as Blackstone, and can also provide their context. Field collection on a systematic basis in the Severn Valley in future may well yield key lithic assemblages for comparative study, such as have been collected and studied elsewhere since 1978 (Woodward 1978; Woodward 1991, and others). However, to date, little attention has been applied to the assessment and detailed analyses of material composition in large field-collected assemblages.
© Internet Archaeology/Author(s)
URL: http://intarch.ac.uk/journal/issue28/3/4.1.html
Last updated: Wed July 21 2010