Click here to see how to contribute
Stage 1 of the East Midlands Archaeological Research Framework involved the production of five papers assessing the Mesolithic resource for Nottinghamshire (Bishop 2000a), Leicestershire and Rutland (Knox 2000), Lincolnshire (Membery 2000c), Derbyshire (Myers 1999b) and Northamptonshire (Phillips 2000). For Stage 2, the present paper draws upon the county-based assessments and seeks to provide a coherent synthesis of the Mesolithic archaeological resource for the region. The regional data will be assessed in the light of national research issues and frameworks with a view to identifying a regional research agenda for the Mesolithic period.
The region embraces considerable variation in topography, geology, soils, drainage, demographics, historical and contemporary landuse patterns. In discussing the Mesolithic resource for such a diverse, contemporary and artificial construct as the East Midlands we need to be aware that the archaeology relates to groups whose patterns of subsistence, settlement and social interaction almost certainly involved mobility and relations well beyond the boundaries of the region. Consequently, any assessment of the Mesolithic resource in the East Midlands must also consider the wider geographical context.
The East Midlands occupies an interesting geographical position in the history of research on the Mesolithic. The north-west of the region includes the southern end of the Pennine upland gritstone environments that have provided an important and enduring focus for studies of the Mesolithic and in national discussions of the period. Numerous excavations of Mesolithic sites, mainly just to the north, in South and West Yorkshire (Radley and Marshall 1963; 1965; Radley and Mellars 1964; Radley 1969; Jacobi et al. 1976), have established the importance of this upland archaeology and ensured a continuity of academic interest (Jacobi 1978a; Jacobi et al. 1976; Myers 1989a; Spikins 1999; Williams 1985). To the south and east lie the Eastern Counties and the South-East regions. In the Eastern Counties, academic interest into the Mesolithic also has a long established history, benefiting through having Cambridge University located within its boundaries. Important publications of early excavations (Clark et al. 1935; Clark 1955; Warren et al. 1934) and the more recent production of county and regional overviews of the archaeology (Jacobi 1980; 1984; 1996; Smith et al. 1989; Tilley 1979) have reinforced a tradition of research into the Mesolithic period. The Eastern Counties region has also seen the recent publication of an archaeological resource assessment (Glazebrook 1997) and research agenda and strategy (Brown and Glazebrook 2000) that includes a section dealing with Palaeolithic and Mesolithic archaeology (Austin 2000). In the South-East region a similarly long established history of research interest can be identified in the various published syntheses of the period specific to the area (Clark 1933; Clark and Rankine 1939; Jacobi 1978b; 1982; Rankine 1949).
To the west lies the West Midlands region. Here too, the Mesolithic received relatively little research interest prior to the 1970s. Much of what had been discovered and published came from incidental finds during field investigations of later prehistoric sites (Gunstone 1962; 1964; 1965a and b). Even when found, Mesolithic material was not always immediately recognised as such (Saville 1972-3). In the 1970s, cave excavations in north Staffordshire provided important Mesolithic evidence (Kelly 1976). More recently fundamental typological work on lithic assemblages has been undertaken (Saville 1981a). Yet in spite of this, as far as the Mesolithic is concerned, the West Midlands remains one of the most under-researched parts of England.
From this it can be appreciated that the East Midlands sits astride historical fault-lines dividing areas with very different Mesolithic research histories. Some areas have well-established traditions of fieldwork and research interest whilst others do not. Developing a regional research agenda will hopefully provide a stimulus for Mesolithic research across a wider area of central England.
The Mesolithic covers the period from the end of the Devensian glaciation, c. 10000 BP, to the first appearance in our archaeological record of what we recognise as the Neolithic, c. 5500 BP (unless otherwise indicated, dates are given as radiocarbon years BP). The intervening millennia were a time of dramatic environmental change in Britain. The end of the ice age saw a rapid warming of the climate. Climatic amelioration initiated widespread successional changes in vegetation patterns, with open late glacial environments being replaced by forests of birch and pine that in turn gave way to more thermophilous species such as oak, elm, and lime. The forests of the postglacial attracted a range of suitably adapted animal species – red deer, roe deer, aurochs, boar and elk – replacing the horse, arctic hare and reindeer that had populated the more open late-glacial landscapes. Sea levels began to rise, gradually inundating the extensive low lying plains, wetlands, lakes, coasts, river valleys and estuaries of the North Sea basin and other coastal areas culminating with the insularisation of Britain. Throughout this period of major environmental change the populations of the Mesolithic pursued their lives and made their living through the hunting of animals, fishing and gathering plant foods. At around 5500 BP, the material culture associated with the Mesolithic disappears from the archaeological record to be succeeded by that of the earlier Neolithic. Apart from new types of stone tool there is the first use of pottery, of the construction and use of burial mounds and the first undeniable evidence for the exploitation of domesticated animals and plants.
The archaeological record of the Mesolithic period in England is dominated by lithic technology. The necessary conditions for the preservation of organic cultural materials or food remains have only rarely been encountered on Mesolithic sites (Clark 1954; Mellars and Dark 1998; Wymer 1962). Human skeletal remains are virtually unknown (but see below). Evidence for structural features only occasionally survives (Clark 1954; Higgs 1959; Radley and Mellars 1964), and then is often limited (Radley et al. 1974) or highly ambiguous (Clark 1934; Clark and Rankine 1939). Furthermore, although the list of reliable radiocarbon determinations for the Mesolithic has grown steadily, features preserving carbon from which secure determinations can be obtained are generally found only through a combination of careful excavation and good luck. In contrast, lithic assemblages survive to provide the most durable, widespread and readily recoverable form of evidence for this period.
Current evidence suggests that the extreme cold conditions of the Younger Dryas (LGIII) may have forced late glacial populations to abandon Britain altogether. From around 10300 BP, lithic industries described as ‘long blade’ appear to represent the first recolonisation from the continent. This event was almost certainly achieved in tandem with the remarkably rapid climatic amelioration – perhaps no more than 50 years in duration (Mellars and Dark 1998, 237) – that marked the end of the Younger Dryas. These industries are characterised, as the name suggests, by the production of long (12 cm or more) blades made from very large bipolar cores (Barton 1997), and are sometimes associated with small, stout obliquely blunted points. Some association with the hunting of horse can be claimed (Moore 1954; Clutton-Brock and Noe-Nygaard 1990), and sites have been recognised from a significant number of open air sites. Often thought to represent the final phase of the late Upper Palaeolithic, these industries are poorly understood. The similarity between some long blade lithic forms and those of what is traditionally recognised as the earliest Mesolithic, together with a recognised compression of dates in the relevant part of the calibration curve (Mellars and Dark 1998, 238), have raised issues regarding the relevance of such distinctions (Barton 1997, 1998).However one regards the long blade technology, by about 9700 BP industries that are recognisably Mesolithic had replaced them. The lithic industries of the first Mesolithic characteristically contain a range of large microlith types manufactured using the micro-burin technique. These so-called non-geometric microliths include varieties of obliquely blunted point, isosceles triangles and elongated trapezoids. One characteristic of these types is that their final shapes depend greatly upon the shape and size characteristics of the blades from which they were made. The blades tend to be quite long, with parallel single or double ridges. These earliest Mesolithic industries are also associated with transversely sharpened flint axes/adzes, scrapers made upon the ends of blades, and a range of burins.
At sites such as Star Carr (Clark 1954) and Thatcham (Wymer 1962) the association between these industries and the hunting of a range of forest-adapted species confirms the explicitly post-glacial adaptation that these assemblages of the ninth millennium cal. BC represent. Typologically they have been recognised as belonging to a complex that can be traced across the dry land bridge of the North Sea basin, or ‘Doggerland’ (Coles 1998), to sites in northern France, the Low Countries, northern Germany, Denmark and southern Sweden. Together, these earlier Mesolithic assemblages belong to what has been called the Maglemosian techno-complex (Jacobi 1978a).
In England recent accelerator radiocarbon dating of resin residues adhering to microliths (Roberts et al. 1998) has indicated that assemblages combining obliquely blunted points and elongated trapezoids – in northern England, the so-called ‘Star Carr-type’ (Jacobi 1978a) – may represent the earliest Mesolithic. Other earlier Mesolithic assemblages characterised by the dominance of obliquely blunted points, often exhibiting opposed retouching at the tip, and in which elongated trapezoids are absent – the so-called ‘Deepcar-type’ (ibid.) – may represent a slightly later chronological phase. The latter extend across northern England from the Pennines (Radley and Mellars 1964) to sites in East Yorkshire (Manby 1966), the North Yorkshire Moors and Lincolnshire (Jacobi 1978a), and at least as far south as southern Derbyshire (Manby 1963). Debate concerning the significance of these typological variants amongst earlier Mesolithic assemblages has also raised the possibility that they represent distinct social groupings within the Maglemosian techno-complex (Jacobi 1978a). This debate has been fuelled by observations regarding their spatial distribution and contrasting patterns of raw material use.
Analysis of ‘Deepcar-type’ assemblages (Myers 1986; 1989a) has indicated that large blade cores of Wolds flint, a distinctive opaque mottled grey/white material, were pre-formed at and transported from sites in the Trent valley such as Misterton Carr (Buckland and Dolby 1973) to bases established within the valleys of the Pennines, represented by sites such as Deepcar (Radley and Mellars 1964). Here it is argued that the cores were used to gear-up the technology in anticipation of task-group activity undertaken away from the bases at a series of temporary field camps and hunting sites located at higher altitude. This activity chain is traced through the debitage, by-products and discard patterns of material from excavated sites. The ability to follow raw material reduction sequences from lithic procurement to site discard patterns across the region provides important data that can contribute towards the development of perspectives on the scale and character of earlier Mesolithic mobility.
Discussions on the nature of earlier Mesolithic subsistence and mobility have of course been significantly shaped through the interpretations placed upon the faunal evidence recovered from Star Carr (Caulfield 1978; Clark 1972; Clutton-Brock and Noe-Nygaard 1990; Day 1996; Dumont 1989; Legge and Rowley-Conwy 1988; Mellars and Dark 1998; Pitts 1979). It is interesting to note that analyses of the faunal assemblage have given rise to very different interpretations regarding the seasonality and nature of activity at Star Carr. However, most authors would subscribe to a view of earlier Mesolithic mobility that incorporates sizeable territories within annual movements. Much of the mobility of these communities will have been in the form of task-groups moving from and ultimately returning to a small number of established residential bases. In the process, these task-groups will have created a series of base camps and activity locations (sensu Binford 1978). The integration of upland and lowland environments within a single annual cycle of exploitation is a theme common to most discussions of this period in northern England.
The great unknown regarding these settlement systems is the full extent of activity undertaken within lower lying environments. Heavily alluviated river valleys, such as the Trent, and the inundated lowlands of the continental land bridge fulfilled an uncertain role in the lives of earlier Mesolithic populations of the area. That earlier Mesolithic artefacts have been trawled up from the North Sea (Godwin and Godwin 1933) is well known. It must be suspected that these variously deeply buried or drowned lowland landscapes were critical to the settlement systems, if not for the primary residential sites themselves, of at least some earlier Mesolithic communities with whose archaeology we are engaged. Consequently our field evidence may itself be biased, providing glimpses of only selected parts of the total settlement system.
Ever since Clark’s seminal works (1932; 1933) sub-division of Mesolithic industries on typological grounds into at least two chronological phases has been generally accepted. The increasing numbers of radiocarbon determinations (Jacobi 1973; 1976; Mellars 1974; Myers 1989b) have in general supported this typological sub-division. The typological characteristics of earlier Mesolithic tool assemblages have already been discussed. At around 8650 BP the assemblages characteristic of the Maglemosian techno-complex disappear. In their place assemblages exhibiting different typological and technological characteristics become widespread. The large non-geometric forms of microlith are replaced by a wide variety of smaller, so-called geometric forms. These include scalene and isosceles triangles, varieties of backed rods, small oblique points, rhomboids, crescents and micro-denticulated forms. In contrast to earlier Mesolithic types, the shapes of later Mesolithic microliths generally owe less to the shape and size of the flake/blade from which they were made than to the use of retouching. For certain types of microlith, such as rods, the micro-burin technique for the careful detachment of the bulb of force was not always applied. The tranchet axes/picks and their distinctive re-sharpening flakes that feature in earlier Mesolithic assemblages also virtually disappear after 8650 BP.Although the transition from earlier to later Mesolithic industries does appear to be a widespread and more or less synchronous phenomenon, there is a possibility that in some areas of south-east and central England an intermediate development can be distinguished. In and around the Weald it has long been recognised (Woodcock 1973) that there are assemblages containing a range of distinctive microlith types – hollow-based points, points with inverse basal retouch – alongside other, typically earlier Mesolithic forms. These ‘Horsham’ industries have been recognised as possibly representing a localised intermediate industrial phase between the earlier and later Mesolithic (Jacobi 1978b). Like their chronology, the overall spatial distribution of these distinctive assemblages remains to be firmly established.
Typological research in the Midlands has indicated that assemblages bearing some similarities to those of the ‘Horsham’ industries may be present (Saville 1972-3, 1981a: Phillips 2000). Over how wide an area such assemblages may be found is not known. Certainly, assemblages containing obliquely blunted points with inverse basal retouch have been reported from within the East Midlands. The evidence from sites as far north as West Keal in Lincolnshire (Jacobi 1984, 56) and as far south as Duston, Northamptonshire (ibid.; Saville 1981b) implies that such industries may be found over a large part of the region. Reynier (1994; 1998) has examined the chronological and spatial characteristics of the ‘Star Carr’, ‘Deepcar’, ‘Horsham’ and Midland assemblage varieties. He suggests that it is possible through these assemblages to begin to recognise the gradual infilling of much of the landscape with evidence for Mesolithic activity. In developing a research agenda the possibility that the region contains such typologically and potentially chronologically distinctive industries should certainly be considered (Phillips 2000).
Detailed typological study and cluster analysis of later Mesolithic microlith assemblages in England (Jacobi 1976; 1979) has identified a series of ‘style zones’ based on variations in the representation of differing microlith types. Some of the zones identified are suggested as possibly reflecting social territories within which traditions of microlith manufacture evolved and were sustained. Reynier (1994; 1998) has proposed that the initial formation of these regional territories can be traced back to the earlier Mesolithic assemblage varieties discussed previously. Interestingly, the size of the territories indicated by Jacobi’s analysis appear to be significantly smaller than the areas covered by the Maglemosian sub-types. In northern England, for example, Jacobi’s backed-bladelet grouping is largely confined to upland regions of the Pennines and North Yorkshire. The stylistic evidence that later Mesolithic territory sizes in England may have been smaller than those of the earlier Mesolithic complements similar analyses elsewhere in continental Europe (Gendel 1984, 1987).The function of microliths in Mesolithic assemblages has long excited debate. In response to the long established tradition in British archaeology of regarding them as weapon armatures, Clarke (1976) provided a timely reminder of the diversity of functions to which such composite technological components could be usefully combined and applied. That they may have served a range of functions is not in question. However, there is a great deal of evidence that positively links microliths to hunting weaponry. The changes in microlith styles that characterise the earlier to later Mesolithic transition appear to be associated with an increase in the complexity of Mesolithic projectiles (David 1998; Myers 1989b). It is suggested that after 8650 BP the numbers of armatures being fitted onto projectile shafts increased dramatically. An understanding of the different demands placed upon the scheduling of weaponry manufacture and maintenance by differing hunting strategies may provide some explanation for these observed technological changes (Eerkens 1998; Myers 1989a; b; Zvelebil 1984). Recent research has also begun examine our understanding of alternative strategies in the production of microliths (Finlay 2000).
The changing microlith styles that characterise the earlier to later Mesolithic transition represent just one aspect of a series of observed changes that occur in technology at or around the mid ninth millennium BP. It has long been recognised (Buckley 1924) that the morphology of debitage in Mesolithic assemblages changes in parallel with the changes in microlith styles. In the past this has been misleadingly characterised as being a shift from ‘broad-blade’ (earlier) to ‘narrow-blade’ (later) assemblages. Metrical analyses of site assemblages, predominantly from southern England, have subsequently demonstrated that in comparison with those of the later Mesolithic, assemblages up to c. 8650 BP actually contain a higher proportion of pieces that are longer in relation to breadth (Pitts and Jacobi 1979).The change in the morphology of flakes and blades provides a useful approach in lithic analysis for assigning general chronological phases within or between assemblages. Yet it must be acknowledged that until the metrical analysis of dated Mesolithic assemblages has been extended to the Midlands and the North serious doubts must remain concerning the utility of this approach outside southern England. Even so, there is a general consensus that throughout England, earlier Mesolithic assemblages demonstrate a greater emphasis upon the production of longer blades with parallel margins and straight dorsal ridges.
In many regions of Britain it has also been observed that varieties of stone being exploited for the lithic technology also change (Jacobi 1978a; 1979; Radley and Mellars 1964) during the mid ninth millennium BP. From south-western (Jacobi 1979; Pitts and Jacobi 1979), southern (Care 1982; Pitts and Jacobi 1979) and northern England (Jacobi 1978a; Myers 1989a; Radley and Mellars 1964) it has been reported that the earlier to later Mesolithic transition coincides with a change in the balance of lithic raw materials being exploited in assemblages. In the south a heavy dependence upon chalk flint is reduced through a pronounced increase in the use of low-grade river gravel flint and Greensand cherts. The South-West sees high-grade translucent flint and Cretaceous flint from eastern sources supplemented or replaced in later Mesolithic assemblages by low-grade beach flint and various cherts. A similar pattern of change has been identified amongst Pennine assemblages where, after 8650 BP, assemblages dominated by cherts obtained from the various Pennine limestones are found throughout the area.It is hard to generalise satisfactorily about the nature of the raw material changes that take place during the mid ninth millennium BP. Higher quality materials continue to be used. They are, however, supplemented and locally replaced by a range of lower grade materials. The area over which any particular raw material is used appears to relate more closely to the procurement source after 8650 BP. This observed change in raw material distributions may bear some relationship to the changing sizes of territories exploited by Mesolithic populations (Spikins 1999, 10). In areas where better quality materials do not occur naturally this may have prompted an increased reliance upon lower grade materials. These chronological shifts in raw material procurement may also relate to changes in the scheduling of technological activity (Myers 1989a; b).
Observations on Mesolithic raw material use and distribution patterns have, over many years, stimulated discussion regarding the mechanisms responsible. As already discussed with regard to the use of Wolds flint in earlier Mesolithic assemblages in northern England, it is possible through the analysis of reduction sequences to look in more detail at the structure and movement of raw materials and their use across regions. As yet little detailed work of this kind has been done on later Mesolithic assemblages. However, it has been suggested that the use of black Derbyshire chert during the later Mesolithic may involve the transportation of relatively unmodified nodules or lenticular slabs away from the sources (Myers 1986). Evidence for the initial setting-up and use of cores appears to be represented at many sites in differing locations and at widely varying distances from the source area.Although much more detailed work is needed in differing regions it is tentatively suggested that the scheduling of later Mesolithic technological activity was designed to meet more short-term goals. This would contrast with the planned and structured procurement and reduction sequence discussed for earlier Mesolithic assemblages. It would be of great interest to know if other researchers had established through the refitting of later Mesolithic site assemblages the condition of the nodule/core prior to reductive activity.
The comparative statistical analysis of Mesolithic tool assemblages has received remarkably little attention over the years. Mellars'(1976a) study remains the only substantive, large-scale attempt to define site-types primarily on the basis of variations in the representation of a range of tool forms. This important work was based upon the quantitative analysis and comparison of assemblages using a restricted number of recognised tool types. The analysis compared Mesolithic assemblages without reference to their typological or chronological position within the period.The analysis produced a series of assemblage types ranging from ‘balanced assemblages’, in which a wide range of the tool types were represented, through to more specialised assemblages where individual tool types predominated and the range of tool types represented was restricted. Amongst the balanced assemblages were a distinct grouping of sites from the Pennines that are all earlier Mesolithic in date. Amongst specialised assemblages were a small number in which scrapers dominated. A much larger number of sites classified as ‘microlith-dominated’ were also recognised. Significantly, the majority of the latter were also from upland excavations and proved to be later Mesolithic in date. It has been suggested that the increase in the complexity of composite weapons after 8650 BP may be substantially responsible for the statistical grouping of earlier Mesolithic balanced assemblages and later Mesolithic microlith-dominated sites (Myers 1987) in this analysis.
Observations on the nature of the earlier to later Mesolithic transition have also been made regarding non-technological characteristics. In particular, contrasting spatial patterns in earlier and later Mesolithic settlement evidence have been recognised. A number of authors have identified a tendency for later Mesolithic evidence to be both more widespread, occupying a greater variety of environments, and more numerous at a regional (Jacobi 1973; 1979; Morrison 1980, 136; Shennan 1985, 49) and more local scale (Myers 1986, 226; Spikins 1999; Williams 1985, 17). Drawing attention to the evidence from well-surveyed areas, Jacobi noted the predominance of later Mesolithic sites in the Cotwolds, the Weald and Kent, Surrey and Sussex, and in the upland Pennines and Clevelands. Later Mesolithic distributional evidence is ‘in sharp contrast to that gained for the earlier period where sites are both fewer… and appear restricted to a limited area’ (1973, 247). Jacobi (1976) calculated a six-fold increase in later assemblages over those of the earlier Mesolithic (Spikins 1999, 8). Manby’s (1963, 11) survey of Mesolithic settlement evidence across South and West Yorkshire, Derbyshire, Nottinghamshire and Lincolnshire indicates a similar preponderance of typologically later assemblages.Others have also noted that in comparison with earlier periods the later Mesolithic sees a general reduction in site size (Care 1982; Myers 1986). While the earlier Mesolithic also has small sites, such as those along Mickleden Edge (Radley and Marshall 1965), there appears to be a significant increase in the representation of such small sites during the later Mesolithic, both in terms of area and assemblage size (Mellars 1976a; Myers 1987). It must be acknowledged, however, that the analysis and interpretation of site sizes, either in terms of area or quantities of material is notoriously complex (Mellars 1976a, 377-8).
In characterising the structure of settlement patterns across a region it will be necessary to look at both the spatial and the compositional elements of lithic scatters (Spikins 2000). Characterising assemblages through the analysis of tool assemblages may represent one possible approach, but the effects of changes in composite tool complexity will need to be considered from the outset. Separate analyses of at least the earlier and later Mesolithic would seem justified. Research that combines the consideration of inter-assemblage variability in both tools and the reduction evidence provided by the debitage provides a more broadly based approach that can shed light upon the dynamics of lithic reduction and tool manufacture/discard patterning.
To summarise, widespread changes in the archaeological record have been identified that coincide with the earlier to later Mesolithic transition at around 8650 BP. In some areas an intermediate technological development may be recognisable. The changes cover aspects of tool typology and design, assemblage content, debitage and reduction sequences, and raw material procurement and distribution. Patterns of settlement also appear to reflect changes at this time. The number of sites increases significantly, with a wider range of environmental locations producing evidence for activity than previously. The increase in site numbers may be associated with a higher representation of smaller sites in the archaeological record. The development of regionally defined styles in microlithic assemblages may be traced from the earlier to later Mesolithic. Such style zones appear to become smaller in area through time. Understanding the significance of these changes in the form, content and distribution of material cultural must be an important consideration in the development of a research agenda for the period, and is reflected in the national research priorities identified by the Prehistoric Society Working Party (1999).
Despite the widespread and far reaching nature of the changes that take place in the archaeological record at around 8650 BP, surprisingly little attention has been given to examining why these changes occur and what they may signify. The comparative absence of faunal and floral data for much of England has not entirely discouraged attempts at modelling later Mesolithic economy. Yet the lack of direct economic data has created considerable room for uncertainty and debate. Views on the nature of Mesolithic economy in England have diverged significantly with some authors contemplating a heavy dependence upon plant foods (Clarke 1976; Jacobi 1978c; Wymer 1991). Others remain convinced that hunting and fishing provided the essential basis for Mesolithic economy (Myers 1986; Smith 1992a). The potential role of coastal resources has also attracted a great deal of discussion (Bonsall 1981; Jacobi 1979). Yet few have attempted to account for the changes in the technology and settlement patterns of Mesolithic groups during the mid ninth millennium BP.Myers (1989b) has argued that changes in the spatial and temporal structure of environmental resources, and particularly game species, during the ninth millennium BP saw a shift in hunting strategies. It is argued that in place of a mixed strategy of logistical and encounter hunting operated during the earlier Mesolithic, strategies became almost wholly encounter-based. The implications for technological structure and organisation are considered alongside those for changing patterns of settlement. Spikins (1999) has considered the spatial evidence for Mesolithic activity against detailed supra-regional modelling of environmental, and specifically vegetational change. Her approach has placed the mid ninth millennium BP transition against the backdrop of a long-term increase in population density, increasing territoriality and competition for resources. It is perhaps worth noting that while these two models provide different perspectives, they are not mutually exclusive.
Environmental evidence, in the form of pollen sequences or charcoal, for the apparent disturbance of Mesolithic forest ecosystems and the creation and/or maintenance of clearings through the use of fire has been widely reported (Jacobi et al. 1976; Jones 1976; Simmons 1964; 1969; 1975; Smith 1970; Turner and Hodgson 1983; Williams 1985). Although the possibility of identifying such evidence in earlier Mesolithic contexts has been discussed (Bush 1988) the overwhelming majority of available evidence indicates such activity after c. 8650 BP. Indeed, a number of authors have pointed out that the evidence for forest burning increases dramatically after c. 7000 BP (Myers 1989a, 135; Simmons and Innes 1987, 389; Spikins 1999).The regeneration of cleared or burned areas of forest promotes, through increased floristic productivity and diversity, conditions that benefit and attract game. The significance and potential benefit of such deliberate forest burning strategies to hunter-gatherer populations has been discussed in detail (Mellars 1975; 1976b). That such strategies may have become a widespread component of later Mesolithic activity is an interesting and potentially significant consideration. Clearly, evidence that might contribute towards an understanding of the chronology, nature and extent of this activity is potentially of great value. Why does the evidence for such strategies increase, particularly after 7000 BP? Archaeologists should consider what impact upon the spatial distribution of activity the creation and maintenance of such clearings would have had over time. Are such patterns detectable in the known or the recoverable archaeological record?
Our radiocarbon evidence for the critical few hundred years when the later Mesolithic ends and the early Neolithic appears is patchy. It is generally accepted, as a ‘rule of thumb’, that around 5500 BP across England and Wales the industries of the later Mesolithic disappear. In terms of material culture, of course, the early Neolithic brings the first appearance of pottery in the archaeological record. The core and bladelet industries and complex composite weaponry of the Late Mesolithic, are replaced by flake core industries and a range of distinctive, bifacially thinned and shaped arrowheads. Yet earlier Neolithic flint industries across southern England also can exhibit a relatively high proportion of bladed pieces (Pitts and Jacobi 1979). Archaeological survey may demonstrate a degree of spatial coincidence between later Mesolithic and early Neolithic industries. In his analysis of data from north-east England, Young (1989, 167-8) has detailed the interpretative problems with such patterning. The chronology of the replacement of later Mesolithic industries by those of the early Neolithic is far from clear. The detail of this missing chronology represents perhaps the most important and archaeologically recoverable data that we could obtain in advancing our efforts to understand the significance of this cultural change mid-way through the sixth millennium BP.Beyond chronology, however, there are fundamental conceptual problems to be overcome. Research looking at the character of hunter-gatherer/farmer contacts (Zvelebil and Rowley-Conwy 1984; Zvelebil 1986) provides useful perspectives that may have some general relevance in describing what happens at frontiers between farmers and hunter-gatherers. Archaeologists dealing with the later Mesolithic-early Neolithic transition have found some utility in the three defined phases used to describe the processes of hunter-gather and farmer contacts: availability, substitution and consolidation (Young 1989). Yet despite their conceptual clarity they do not overcome the uncertainty that exists over the very character of the frontier between later Mesolithic and early Neolithic cultures in England.
On the one hand there is a body of thought that looks for traits traditionally associated with the Neolithic in late Mesolithic contexts. Evidence for the use of recognisably domesticated resources by technologically later Mesolithic groups has been sought, but sites of this period have thus far offered infamously poor conditions for the physical preservation of faunal and floral evidence. Pollen spectra have also failed to provide unambiguous evidence for the cultivation of crops by technologically later Mesolithic groups. There is also a singular absence of tools that can be unambiguously assigned to vegetable food processing – mortars, pounders or grinders – and of structural evidence that might suggest the bulk storage that could accompany an emphasis on plant foods. Yet interest in the possible role of plant foods in the economies of the late Mesolithic and their significance in the transition to the Neolithic has grown in recent years (Zvelebil 1994). The issues surrounding selective forest burning and the manipulation of game populations have also raised the possibility of discussing issues of intensification and the control of animal populations. Yet despite the potential for discussing early moves towards domestication, Mesolithic sites and assemblages of the early sixth millennium BP, where recognised through radiocarbon dating, still look very much akin to those dating to just after 8650 BP.
On the other hand there is a long-running, fundamental debate concerning the very nature of early Neolithic society in Britain (Bradley 1984; Thomas 1991; Whittle 1988). Whereas Neolithic societies were once assumed to have been sedentary exploiters of domesticated plants and animals, archaeologists are now invited to consider early Neolithic society not only in terms of pots, leaf arrowheads, agriculture, cattle and collective burial but also mobility, hunting, and seasonality -themes more traditionally at home in hunter-gatherer research.
In discarding old assumptions about the character of late Mesolithic and early Neolithic communities archaeology has stimulated renewed interest in the significance of the changes in material culture that take place in the mid sixth millennium BP. It is no surprise therefore that in the face of such uncertainty the transition from the late Mesolithic to early Neolithic in Britain should be one of the research priorities identified by the Prehistoric Society (1999).
Having presented the background it is possible to evaluate the Mesolithic resource in the East Midlands (Fig. 18). A broadly historical consideration of the state of the resource will assist in placing the current position in a developmental context. In this way it is hoped that historical bias in fieldwork and research traditions may be overcome. Using the county assessments as a starting point the rest of this paper will seek to identify how fieldwork and other research in the region can make the greatest contribution towards the wider study of the Mesolithic period.
The county assessments confirm that the character of the Mesolithic archaeological record largely mirrors the situation found elsewhere: organic evidence of diet or technology is virtually absent, structural evidence is very limited, numbers of secure radiocarbon determinations are few, and lithic findspots and scatters provide the overwhelming preponderance of the available evidence. In short, the archaeology consists of a relatively small number of excavated sites and a great deal of information from surface collections.There is a general consensus within the county assessments that the transitional archaeologies of late glacial-early postglacial hunting and gathering and of the Mesolithic to Neolithic are of great importance and must be included. Both of these topics have already been identified as national research priorities by the Prehistoric Society (1999) and are also echoed in the Eastern Counties Research Agenda (Austin 2000; see also French 1992 for Fenland research priorities).
It is however of some concern that the county assessments appear to attach only limited significance to consideration of the earlier and later Mesolithic periods. Indeed, the impression might be gained that the archaeology of this five and a half thousand (calendar)years during which the environment of Britain changed so fundamentally is to be handled as a single undifferentiated phenomenon. Yet it must be at least suspected that the adaptations of the first thousand years of the postglacial are not necessarily the same as those of the sixth millennium BP. This raises a fundamental issue concerning the SMRs and Mesolithic research. Does the lack of such a discussion reflect a general absence of chrono-typological detail in SMR coverage of the Mesolithic?
SMRs may be forgiven for a lack of chronological or typological refinement in their Mesolithic data when it is remembered that the Gazetteer of Mesolithic Sites (Wymer 1977) also offered no chronological refinement. The gazetteer will have been the primary source for Mesolithic data for many SMRs when they were first established during the 1980s. Nonetheless, the absence of any such distinction, either in the records and/or in the assessment of those records carries important implications for the development of the East Midlands research agenda. All agree the importance of including the key transitional issues of the lateglacial to early postglacial adaptations and of the Mesolithic to Neolithic. Yet the lack of detail in their records raises serious questions concerning the utility of SMRs for addressing such important issues. Furthermore, it denies any research utility for the SMR in addressing a third topic identified by the Prehistoric Society (1999) as a research priority, but left largely unexplored by the county assessments and completely left out of the Eastern Counties research agenda – namely that of identifying and understanding continuity and/or change during the Mesolithic.
The county assessments illustrate the contrasting historical backgrounds to research across the region. They show how the differing research environments, compounded by differing patterns of land use and opportunities for site recognition have influenced the development of the known Mesolithic resource in the East Midlands. They also emphasise the importance of environmental evidence in developing an understanding of the Mesolithic across the region. Whilst highlighting the relatively small number of sites that have actually produced relevant environmental evidence, all acknowledge that palaeoenvironmental information is essential in understanding the character of Mesolithic activity. Attention is focussed upon the potential that exists across the region for the preservation and recovery of such evidence.
Early investigations Up until the 1960s little was known about the Mesolithic for most of the East Midlands. The generally small number of pre-1960s published accounts reporting, often incidentally, on the presence of Mesolithic material from Northamptonshire, Leicestershire, Lincolnshire and Nottinghamshire (Barley 1959; Bowen 1945-47; Browne 1888; Dudley 1949; Everard 1946; Hunt 1908; Posnansky 1955a; 1956a; Pickering 1918; Rankine 1951) suggests that the known archaeological resource was extremely limited.
The main exception was to be found in the north of the region. The caves in the Magnesian limestone of north-eastern Derbyshire bordering with Nottinghamshire, and the Carboniferous limestone of the White Peak had attracted archaeological interest from the 1800s onwards. The caves provided archaeologists with obvious targets for investigation by excavation. Although primarily focussed upon the potential for discovering Palaeolithic remains, a number of cave excavations provided evidence of Mesolithic activity. On the Magnesian limestone, excavations at Mother Grundy’s Parlour (Armstrong 1925), Pin Hole Cave (Armstrong 1926, 1929a; 1929b; 1937), Yew Tree Shelter (Armstrong 1938), Ash Tree Cave (Armstrong 1956; 1957) and the Whaley rock shelters (Radley 1967) all yielded Mesolithic material (Fig. 19). Similar if rather ephemeral discoveries were made in the Carboniferous limestone caves of the upper Dove valley at Dowel Cave (Bramwell 1959) and Foxhole Cave (Bramwell 1971a; 1971b; 1972; Jackson 1952).
The long-established tradition of flint collecting from erosion patches in the upland peat moorlands of the Pennines around Huddersfield gave rise to some of the earliest published regional syntheses of Mesolithic material in England (Buckley 1924; Petch 1924). The erosion of peat on the upland moors of northern Derbyshire (Radley 1962) combined with improved public access to the moors prompted a growth of interest in flint collecting and an increased awareness of Mesolithic evidence in the area (Radley 1963a). Selected parts of northern Lincolnshire (Armstrong 1931; 1932a; 1932b; Dudley 1949; Gatty 1900-1) also had a long-established history for unsystematic flint collection from eroding cover sands and peat.
Garrod’s The Upper Palaeolithic Age in Britain (1926) and Clark’s The Mesolithic Age in Britain (1932) represented the first national syntheses of the archaeology of these early periods. They were instrumental in establishing the wider academic context within which the research agendas of contemporary and subsequent generations of archaeologists were to develop. They drew heavily upon the cave excavations in northern Derbyshire, particularly those at Creswell. They also made known to a wider academic audience the Mesolithic discoveries in the Pennines and Lincolnshire. In particular, Clark was able to make use of information from the Huddersfield and Marsden districts obtained through Francis Buckley’s excavations, collections and related publications (Buckley 1924; Petch 1924).
Once established in such national period discussions, subsequent researchers have been drawn back to the same areas to re-examine existing collections or undertake new fieldwork. In this sense, the known Mesolithic resource of these areas has benefited historically through their early recognition. Other parts of the region have also seen early examples of fieldwalking collections, such as the extensive assemblage from fields around Duston, Northamptonshire (Phillips 2000). Yet such work has only recently been drawn into regional and national syntheses. Consequently these areas have not attracted further field research and have been slow to impact upon perceptions regarding the importance of Mesolithic archaeology in the region.
The resulting historical bias in research interest and fieldwork encouraged a widely held impression in the region that Mesolithic fieldwork was very largely a phenomenon of caves and upland areas. Outside such locations Mesolithic material tended to come to light only by chance during the investigation of more recent archaeological sites. The paucity of pre-1960s published accounts for the southern half of Derbyshire (Lomas 1959) also reflects this bias in the history of fieldwork and research interests, and is more akin to the situation elsewhere in the East Midlands. For the vast majority of the region it is fair to say that, as with the West Midlands, the Mesolithic has historically received little attention. The consequences for the discussion of research priorities within the East Midlands have been all too obvious. The last published attempt to define research priorities for the East Midlands (Mahany 1977a) illustrates that even as late as the 1970s Mesolithic research was assigned a low priority. Even today the region does not appear to contribute substantively to the profile of Mesolithic research (Young 2000) in England.
1960s- to the present
The bias that developed in fieldwork during the first half of the last century resulted in a lack of attention for the Mesolithic of lowland areas of the East Midlands. In practical terms this has meant that for much of the region little concern for Mesolithic archaeology has been demonstrated. Since the 1960s, however, a series of developments have seen new patterns of fieldwork arise.
The discoveries made at Star Carr (Clark 1954) in the Vale of Pickering, East Yorkshire, represented a major turning point. The chance discovery of a site providing extraordinary conditions of organic preservation did much more than broaden our knowledge of Mesolithic diet and organic technology. The evidence recovered from the site has fuelled debate and differing interpretations ever since (Caulfield 1978; Clutton-Brock and Noe-Nygaard 1990; Clark 1972; Day 1996; Dumont 1989; Legge and Rowley-Conwy 1988; Mellars and Dark 1998; Pitts 1979; Wheeler 1978). Crucially, the discovery of such an important site in a lowland location represented a major stimulus for archaeologists to begin actively searching for the Mesolithic in similar, previously under-researched localities (Radley 1968; Wymer 1962).
Manby’s (1963) review of Mesolithic evidence from the Peak District and Trent basin represented an important step in encouraging data collection from parts of the landscape other than caves and the gritstone uplands. His paper identified a more widespread distribution of Mesolithic activity across Derbyshire, Nottinghamshire and Lincolnshire. He drew attention to the small but increasing body of evidence for open air sites away from the gritstones. Specifically, Manby discussed the evidence from a number of open air sites identified on the Carboniferous limestone of northern Derbyshire, and in lower lying areas such as the Coal Measures and the Trent valley in the south of the county.
This rethinking of Mesolithic distributional evidence was instrumental in encouraging the collection of data from open air locations. Those parts of the landscape that had previously been largely ignored and which offered the potential for good preservation became a target for research. In 1966 the attention of Doncaster Museum was drawn to a collection of flints made by a local farmer from eroding lowland fen peats and sandy soils between the Trent and the Idle. Subsequent annual and more systematic collections led to a limited excavation in 1971. This work produced evidence for a substantial earlier Mesolithic assemblage at Misterton Carr (Buckland and Dolby 1973). Although the excavation did not reveal any structural evidence or organic preservation, the assemblage provided a rare opportunity for studying the character of earlier Mesolithic activity in the Trent valley, and has subsequently been incorporated into many discussions of the period across northern England (Jacobi 1978a; Myers 1989a; Spikins 1999).
This general reappraisal of the Mesolithic and fieldwork in low-lying regions did not result, however, in traditional patterns of fieldwork being abandoned. In the south Pennines surface collection was supplemented by a series of very important excavations of Mesolithic sites (Radley and Marshall 1963; 1965; Radley and Mellars 1964; Radley et al. 1974; Stonehouse 1986), including the small site at Red Ratcher (Stonehouse 1976) located at 500 m OD in the very north of Derbyshire. Buckley’s 1920s and 1930s excavations also continued to provide a valuable resource from which new information was gleaned (Radley and Mellars 1964). Radiocarbon dates were obtained from excavated charcoal that had been kept and placed by Buckley in small glass boxes (Switsur and Jacobi 1975). Cave and rock shelter excavations in Derbyshire continued to provide new Mesolithic evidence, as at Foxhole Cave (Bramwell 1971a; 1971b; 1972), the rock-fissure site at Sheldon (Radley 1968) and the so-called rock shelter site at Roystone Rocks (Myers 1992). In the moorland environments of Pennine Derbyshire flint collecting continued to identify new sites (Henderson 1973; Pierpoint and Hart 1980).
The southern Pennine peat deposits also provided the basis for early work on the reconstruction of post-glacial vegetational patterns through pollen analysis (Tallis 1964). Such work continued with the additional analytical benefit of associated radiocarbon dating of pollen sequences (Tallis and Switsur 1973). At about the same time evidence for the potential role of Mesolithic activity in modifying vegetation patterns through the use of fire was being recognised in upland northern Derbyshire (Hicks 1972). As such evidence became more widely recognised the southern Pennines became a focus for more detailed research examining the role of clearance in Mesolithic activity (Williams 1985).
Throughout the East Midlands, Mesolithic evidence continued to come to light through excavations of later prehistoric and historic sites. In Northamptonshire stratified Mesolithic material was recognised during the Ministry of Public Building and Works excavations at Thrapston ironstone quarry, Aldwincle (Jackson 1976a; 1977). Similar rescue excavations within Northampton at Chalk Lane identified a series of intersecting gullies cut into the gravel terrace and containing earlier Mesolithic material (Williams and Shaw 1981). Excavations of Neolithic sites at Briar Hill (Phillips 2000) and Ecton (Moore and Williams 1975) also produced small quantities of Mesolithic material. Limited excavations were also undertaken on the site of a Mesolithic scatter at Beadle Quarry, East Goscote, Leicestershire (Saville 1976).
Occasionally, such chance encounters produced remarkable evidence. In Derbyshire, a trial trench was excavated by Trent and Peak Archaeological Trust during 1984, at a site called Lismore Fields, Buxton, in advance of a housing development. Sitting at a point where the Wye valley widens to form a bowl, Buxton lies at 300 m OD on the junction of the Carboniferous limestone and the millstone grits. Lismore Fields, an area of (then) surviving pasture within Buxton’s south-western suburbs, occupies a low, 175 m wide plateau between two tributaries of the River Wye. The soils of the plateau are capped by a Head deposit, derived from the gritstones, and would have been heavy and wet before the insertion of field drains in the post-medieval period (Garton 1991, 13). The trench was excavated in the belief that a Roman road crossed the area. Instead, four seasons of excavation revealed a later Mesolithic flint industry, possibly associated with a ring-slot and post structure, and an earlier Neolithic settlement with evidence for three rectangular houses (Garton 1987a; 1991, 12-13). A posthole associated with the ring-slot produced a date on charcoal of 5270 ± 100 BP (OxA-2433). The excavation report has not yet been published.
Environmental evidence from Lismore Fields has also made a contribution to the debate on the manipulation of the environment through deliberate burning of forests. Wiltshire and Edwards (1993) examined a pollen sequence that included evidence contemporary with the Mesolithic activity. The presence of charcoal in the sediments suggested that fire had been used, but the pollen analysis indicated localised clearance impacts of relatively small scale and duration.
In terms of extending our knowledge of Mesolithic activity and settlement across the East Midlands, the most significant development has been the increased and more widespread use of surface collection from ploughed fields. Of course, as an approach to data gathering, surface collection has assisted the study of all prehistoric and early historic periods. Yet when comparing Mesolithic structural evidence with that of later periods it appears to be both rare and ephemeral in character. Other techniques for site recognition that benefit the study of later periods, such as aerial photography or geophysical survey, are consequently of little value. Surface collection therefore occupies a particularly important position when prospecting for Mesolithic activity and has been invaluable in overcoming some of the historical biases in our distributional knowledge.
Since the publication of Wymer’s (1977) gazetteer, the total number of sites, be they scatters or individual findspots, has increased from 177 to 753 – an increase of 325%. As Table 1 shows, even where the increase has been lowest, in Nottinghamshire and Lincolnshire, there has still been a significant growth in the number of recorded sites. For Leicestershire and Derbyshire the increase has been greater than 400%. The numbers of recorded sites reflected in the records held by SMRs within the region have increased largely in response to the greater emphasis placed upon systematic surface collection as a fieldwork technique since the 1970s. The development of the systematic and controlled collection of worked lithics from ploughed fields during the 1960s (Radley and Cooper 1968) has been followed by the routine incorporation of this technique into research designs and field evaluation strategies. Programmes of surface collection in Derbyshire (Barnatt et al. forthcoming; Evison 1988; Gerrish 1982; Garton and Beswick 1983; Hart 1981; Knight et al. 1998; Phillips and Guirr 1985), Leicestershire (Liddle and Hartley 1995; Liddle and Knox 1991; 1997), Nottinghamshire (Phillips and Guirr 1985; TPAT 1992), Lincolnshire (Hayes and Lane 1992) and Northamptonshire (Hall 1985; Martin and Hall 1980) have all been identified in the county resource assessments as being to a greater or lesser extent responsible for these dramatic increases in site numbers.
Table 1: Representation of Mesolithic sites in the East Midlands region (* adjusted to exclude those parishes in North Lincolnshire and North-East Lincolnshire)
|County||No. of sites in Wymer 1977||No. of sites (current)||% change 2977||Area in Sq.km||Area in Sq. km per known site|
|Leicestershire/ Rutland||33||175||+ 430||2548||14.56|
The current average density of sites across the region’s landscape of one per 20.83 km² compares with a figure of one per 88.63 km² for 1977. The present figures can, however, be seen to conceal significant variations between the counties. The highest densities are found in those counties where, judging by the county assessments, systematic surface collection programmes have been most actively pursued. The densities for Northamptonshire, Leicestershire and Derbyshire (see Table 1) confirm the impact and importance of survey in enhancing the numbers of known sites and in producing some of the highest densities of evidence in the region.
The systematic surface collection of worked lithics, supplemented in areas of pasture by systematic test-pitting (Edmonds et al. in prep.; Garton and Kennett 1996; Guilbert et al. 1997; Torrence and Edmonds 1988) is an invaluable method for testing the validity of existing patterns in our distributional knowledge. In particular, such surveys can be used to examine whether patterning in our data relates to Mesolithic activity or to historical patterning in the fieldwork endeavours of archaeologists.
The Derbyshire assessment (Myers 1999b) emphasises the overall historical bias in fieldwork within the county, with certain geologies having been largely ignored by archaeologists. The assessment is still able to draw attention to the varied topographical locations in which Mesolithic evidence has been identified. Whilst prominent or even slightly elevated locations may have attracted activity, areas with heavy soils such as Lismore Fields have also yielded important evidence. In some areas proximity to water may have been a critical factor. In north Derbyshire the work of Hart (1981, 25-6) was designed to look for Mesolithic evidence across the main geological divisions. His work confirmed the presence of Mesolithic sites on ‘all of the geological regions’ (ibid., 25). Similarly, a survey (Barnatt et al. forthcoming) designed to look at lithic distributions along a transect from the Carboniferous limestone, across the Wye and Derwent River valleys and onto the gritstone East Moors of northern Derbyshire has established the presence of Mesolithic material in all major zones including floodplain locations.
For Nottinghamshire, Bishop (2000a) draws attention to the impact of survey along the line of the A46 (TPAT 1992), which identified sites within the Wolds, an area previously thought to be devoid of Mesolithic evidence. Membery (2000c) notes the quantitatively small but highly significant impact intensive fieldwalking along the Fen edge in Lincolnshire (Hayes and Lane 1992) has had upon known Mesolithic distributions. He draws specific attention to a series of sites identified in lower lying areas despite a widespread coverage of alluvium, and a number located on ridges and sand isles at places such as Spilsby, Dogdyke, and East and West Keal.
In Leicestershire a series of intensive surveys have been undertaken with the assistance of volunteer groups since the mid-1970s. The heavy clay soils that cover so much of the county have traditionally been dismissed as having little potential for locating prehistoric settlement and consequently attracted little fieldwork (Clay 1989, 111; 2002). Yet the surveys at Lutterworth/Misterton, Brooksby Estate (Liddle and Knox 1991), Grace Dieu priory (Liddle and Hartley 1995), and Medbourne (Liddle 1994) have demonstrated the previously unsuspected extent of prehistoric activity and transformed the known Mesolithic record. One of the strengths of this survey was that it set out to address a range of topographies and soils. In reviewing the results, Knox (2000) states that the Mesolithic evidence appears to be fairly well spread and not restricted to any particular topography or drift geology. Some topographical preferences for site location have, however, been identified. In particular, the Medbourne survey has revealed some preference for Mesolithic site location on prominent ridge ends along the northern bank of the Welland, as well as in small valley bottoms (ibid.).
In Northamptonshire, surface collection surveys (Hall 1985; Martin and Hall 1980; Parry forthcoming) have once again sought to examine the distribution of evidence across different geologies. Phillips (2000) recognises a similar pattern along the flanks of the River Nene to that revealed in the Leicestershire Medbourne survey. Mesolithic activity appears to be preferentially located upon prominent topographical locations on exposed permeable geologies. With reference to the survey work of Hall (1985) and Martin and Hall (1980), he notes that permeable geologies generally appear to have preferentially attracted Mesolithic sites, with clay soils being avoided. Certainly, the Northamptonshire surveys do appear to demonstrate how lithic evidence of all prehistoric periods has been most frequently discovered on light soils (Hall 1985, 31). Phillips also suggests that in Northamptonshire, concentrations of sites along river valleys and in the north-west uplands of the county may relate to the number of rivers that have their headwaters in these uplands and the use of these valleys for movement and exploitation. He suggests that the north-west uplands may have provided an important crossing point between the adjacent river systems.
The impressive quantitative impact that surface collection survey has had upon the known Mesolithic resource of the East Midlands region is undeniable and has also allowed preliminary observations to be made concerning distribution patterns. It would, however, appear that despite the extensive use of surface collection since the 1960s there have been just a handful of instances where large-scale collection survey evidence has been used for targeting the excavation of a Mesolithic scatter.
As a result of pioneering surface collection fieldwork on the Coal Measures, another historically neglected part of the East Midlands landscape, the North Derbyshire Archaeological Trust undertook a rescue excavation of one of the more promising scatters identified along a south- and easterly-facing prominence above the River Drone, north of Chesterfield. The excavations during 1977 and 1978 at Unstone 1 produced some 4066 flints, with the majority coming from an area of c. 120 m2. The assemblage was ‘balanced’ (sensu Mellars 1976a) with microliths, scrapers, awls and burins all represented, along with cores, core maintenance flakes, retouched and unretouched flakes and blades. A series of features, possibly of Mesolithic date were recorded. There were also indications of surviving stratigraphy within the features suggesting at least two phases of occupation. Unfortunately, the excavations at Unstone 1 have yet to be fully published (Ataman 1978; Courtney 1977).
Scatters located through fieldwalking are however sometimes used for locating mechanically excavated trenches in the hope of identifying buried features. Between 1991 and 1995 the Fenland Survey in Lincolnshire identified a concentration of Mesolithic flintwork at Mexican Bridge, Midville. A trenching excavation established that although there was a Mesolithic scatter and an intact prehistoric land surface preserved beneath later alluvium there were no features identifiable (Crowson et al. 2000; Membery 2000c).
The introduction of Planning Policy Guidance note 16 in 1990 (DoE 1990) has probably been the most influential development in promoting the excavation of Mesolithic sites in environments that have historically received little archaeological attention. A similar benefit can be seen in the increasing numbers of investigations of sites across the region presenting a potential for the preservation of environmental data. The critical change has been the introduction of a developer-funded, development-led spatial agenda to archaeological fieldwork. The distribution of developer-funded archaeological fieldwork has little or no relationship to the traditional patterns of fieldwork established in the first half of the twentieth century. In this way development pressure has imposed a new spatial discipline upon archaeological fieldwork. The archaeological evaluation of areas offering nothing but ‘potential’ has become possible throughout the region. Recent evaluations at Lordsmill Street, Chesterfield, in advance of the construction of a motel, uncovered a remarkable example of prehistoric survival in the midst of intensive nineteenth- and twentieth-century development. Originally thought to offer potential for Roman and medieval archaeology the evaluation resulted in the identification and excavation of medieval cess pits and a series of features, possibly tree throws, containing a substantial lithic assemblage dating to the later Mesolithic (Foundations Archaeology 1999). The site occupied a low, pronounced headland on the west side of the River Rother, bounded immediately to the south by the River Hipper. It would have originally have commanded views of the floodplain. The assemblage, dominated by varieties of Derbyshire chert, included a small obliquely blunted microlith, five scrapers, an awl, blade cores, core rejuvenation material, retouched and unretouched flakes and blades. It is only the second excavated Mesolithic site on the entire Derbyshire Coal Measures. As yet the excavation report has not been published, although a report has been lodged with DCC.
In South Derbyshire, evaluation work in the vicinity of a scheduled a Bronze Age barrow cemetery and an Iron Age cropmark complex led to the recognition of an earlier Mesolithic assemblage on Swarkestone Lowes (Elliott and Knight 1999a). The site was located on the eastern end of the narrow east-west ridge of Triassic Mercia Mudstone that forms the Lowes. Here the Lowes are capped by the Etwall sands and gravels, providing improved drainage. The location has commanding views of the surrounding landscape, being about 10 m above the level of the River Trent to the south, the Cuttle Brook to the east and Sinfin Moor to the north. The assemblage, predominantly made from Wolds flint, contains a fairly broad range of knapping evidence, including core maintenance debris, cores, blades, flakes and end scrapers. This is the first excavated Mesolithic site, and one of only a handful of known Mesolithic findspots in the whole of South Derbyshire (Myers 1999b).
In Northamptonshire developer-funded archaeology has seen the excavation of small quantities of Mesolithic material at Brixworth (Ford 1994; 1995; Jackson 1990) and at Towcester Meadow (Walker 1992). The latter site has demonstrated the high potential for site survival in alluviated deposits within low energy streambeds and river valleys (Phillips 2000). At Burton Latimer (Foundations Archaeology 2000) deposits preserving evidence for forest clearance that is thought to be anthropogenic in origin, but containing no cultural materials, have been radiocarbon dated to 5910 ± 40 BP. In Leicestershire an evaluation of land proposed for development adjacent to Thurlaston Brook at Croft Quarry was undertaken in 1993 (Cooper 1993). Located at 67 m OD the site lay just upstream of the confluence with the River Soar on a geology of Mercian mudstone and glacial boulder clay with some sands and gravels. The evaluation and subsequent excavation (Hughes and Roseff 1995) were hampered by de-watering problems in the trenches, but managed to identify a series of pit-like features containing varying quantities of charcoal flecks, and two shallow curvilinear features thought to represent palisade gullies associated with post-ring roundhouses. The small lithic assemblage may be Mesolithic or Neolithic and included pieces from three of the pit-like features. Unfortunately, the excavation difficulties severely limited the extent of the sampling of deposits in the features. The Croft site has however provided environmental samples for analyses of pollen, macro-flora and beetles. Preliminary investigation of the pollen indicates undisturbed forest environments dating to the Atlantic period (ibid., 103). Information from sites like Croft is adding to the range of Mesolithic environmental data available from other Leicestershire sites such as Narborough (Brown 1999), Austin Friars (Shackley and Hunt 1985) and Birstall (Ripper 1998), and making a significant contribution towards the development of our understanding of post-glacial vegetation patterns in the area.
Whilst acknowledging the benefit that developer-funded archaeology is bringing to Mesolithic and palaeoenvironmental research in the East Midlands, it must also be recognised that such work will remain vulnerable to criticism for being piecemeal until strategies for integrating the results into broader research agendas have been developed. The present initiative will hopefully contribute towards reducing such criticism. Recent research based largely upon developer-funded work on geomorphological and palaeoenvironmental data (Knight and Howard 1994; 2004) provides a good example of the strategic application of developer-funded work in developing our understanding of the complex fluvial landscape history of the Trent and Idle river valleys (Bishop 2000a). Such work not only provides an explicit, detailed model of the development of the environment in these valleys, but also provides a predictive framework for identifying where the potential for the survival of sub-alluvial archaeological and palaeoenvironmental evidence is greatest. Developer-funded archaeology thereby contributes towards the constructive targeting of further PPG16 and other research-funded fieldwork.
In Nottinghamshire the proposed construction of a power station at Staythorpe, south-west of Newark, recently gave rise to a programme of archaeological work (Davies 2001) in an area identified as offering important sub-alluvial potential. Located in the Trent valley on alluvium and river gravels overlying Mercian mudstones, the development required the excavation of two gravel borrow pits. Initial coring confirmed the presence of organic deposits, deeply buried beneath alluvium, providing the potential for the preservation of palaeoenvironmental evidence. Subsequent trenching identified three palaeochannels containing organic evidence of which two proved to be later Mesolithic in date. These have provided a series of four dates on wood and reeds ranging between 6640 ± 60 BP (Beta-142217) and 6040 ± 70 BP (Beta-1442218). Pollen and insect data indicates that, during the latter half of the seventh millennium BP, the area was a mixture of alder, willow and aspen carr, with limited grassland and a background of oak, elm and lime on the adjacent gravel terraces. The later Mesolithic deposits also yielded animal bones for a range of species including roe deer and aurochs. Two of the recovered bones bore clear signs of cut marks.
However, the most remarkable find at Staythorpe was the recovery from one of the later Mesolithic palaeochannels of a well-preserved human femur, probably from an adult female, that has subsequently provided a radiocarbon determination of 6790 ± 40 BP (Beta-144016). At the outset it was noted that in England human remains dating to the Mesolithic are exceptionally rare. Indeed, to my knowledge this is the first demonstrably Mesolithic human bone from an open air location in England, and the first not to come from a coastal or cave location in the United Kingdom.
The Staythorpe femur has been analysed for stable isotopes of carbon and nitrogen (Richards 2001). Stable isotope analyses of bone collagen provides evidence of the protein sources in human diets (Richards 2000) for a period of approximately the last ten years before death. Carbon isotope values (d13C) provide information on the levels of marine (fish, shellfish) versus terrestrial (animal meat, plants) sources of dietary protein. Nitrogen isotope values (d15N) provide data on the relative importance of plant versus animal protein sources. The Staythorpe sample provided a d13C value of -20.4, which is indicative of a diet where all of the protein came from terrestrial sources. This indicates that in the ten years before death this individual spent little if any time at coastal locations consuming marine foods. The d15N value was 9.3, indicating a diet very high in animal protein. Together the results suggest that this individual spent the last ten years of her life obtaining the vast majority of protein from terrestrial animal sources.
The discovery at Staythorpe and the resulting stable isotope data contributes to our perspectives on the character of later Mesolithic diet and mobility. Previous stable isotope data on Mesolithic human remains in Britain has come from caves or shell middens in coastal locations. They have demonstrated an emphasis upon marine sources of dietary protein. The Staythorpe sample provides, for the first time, a clear indication that there were populations whose dietary protein came predominantly from terrestrial sources. Taken together, we have clear dietary evidence for the establishment of varied subsistence strategies by at least the seventh millennium BP. Some groups clearly had an emphasis upon coastal resource exploitation while others appear to have been primarily engaged in exploiting non-coastal resources. The evidence may also provide support for the recognition from the Mesolithic archaeological record for the establishment of spatially discrete or territorial patterns. Furthermore, whatever else the Staythorpe individual had eaten in the decade before her death it does not appear that sources of vegetable protein, such as hazelnuts, played anything but a small and probably seasonal role. The pre dominance of animal protein in her diet emphasises the importance we should attach to evidence for change in Mesolithic hunting strategies. +
The known Mesolithic resource in the East Midlands has, in recent decades, been transformed by changes in field practice and planning guidance. Spatial bias in the historic distribution of archaeological endeavour left large areas of the region uninvestigated. Systematic surface collection from ploughed fields has demonstrated that evidence for Mesolithic activity is widespread. Developer-funded archaeology is beginning to provide excavation detail. The impact has been to transform the assessment of the resource in the region. No longer can discussions of the period comfortably characterise the region as an area that was avoided or only thinly populated (Smith 1992b; Smith and Openshaw 1990). The case for integrating PPG16 fieldwork into wider, more strategic frameworks for research has also been demonstrated. The work at Staythorpe has illustrated how such work can produce nationally important and novel evidence concerning the character of subsistence and mobility in the early seventh millennium BP.This is not to imply, however, that under PPG16 all is well. The East Midlands faces a persistent problem with important resources from excavated sites not being adequately published. The archaeological resource represented by upland caves and rock shelters have frequently received only interim reporting of the various excavations. Curatorial concerns for the preservation by record of many of these sites dictates that there should be a determined effort to conserve excavation records for future research. Where appropriate, projects to preserve and analyse such archives and publish the results should be considered. The problem of non-publication is not confined to cave excavations. The county assessments identify a number of important sites that have yet to be published. Quite apart from any moral issues, raising the Mesolithic research profile of the region will require the publication of such excavations.
The apparent lack of chronological refinement in much of the survey data and/or in SMRs restricts the utility of that data for addressing certain important research issues. The changes in technology which define the earlier to later Mesolithic transition provide a variety of opportunities for the assignment of lithic assemblages to one or other of these phases. Notwithstanding the multiplicity of problems that beset the analysis and interpretation of lithic assemblages recovered as scatters, it is possible through the application of typological, metrical and raw material analyses to obtain some indication of the presence of earlier and/or later Mesolithic activity. Future project briefs should specify the need for an appropriate level of analytical detail in lithic analysis to be provided for SMRs.
The surface collection surveys have, however, given rise to a number of intriguing observations regarding the overall distribution of Mesolithic evidence. Throughout the East Midlands there is an observed tendency for Mesolithic activity to be found on high points, ridges, prominences or headlands. Specific observations on this pattern have been made with regards to the gritstone uplands, Coal Measures and Mercian mudstones of Derbyshire, the Fen edge of Lincolnshire, the River Nene of Northamptonshire and the River Welland in Leicestershire. Some of these locations provide vantage points offering views, vegetation permitting, along deeply incised valleys whilst others look out across extensive areas of relatively flat land.
A similar preference for well-drained locations has also been noted in a number of the county assessments (Membery 2000c; Phillips 2000). That such locations appear preferentially to have attracted Mesolithic activity has long been recognised (Rankine 1949; Mellars and Reinhardt 1978). At the same time, however, surface collection surveys in Derbyshire, Lincolnshire and Leicestershire have demonstrated that Mesolithic activity is spread across the landscape in a wide range of locations and on varied geologies and geomorphologies including those with heavier soils. It is interesting that in Leicestershire the long-held assumption that clay areas would have been avoided in prehistory has been dispelled by survey (Clay 2002). In Derbyshire, the chance discovery at Lismore Fields also points to later Mesolithic and earlier Neolithic settlement, not just activity, on heavy soils. Yet in Northamptonshire extensive surveys of varied geologies and geomorphologies have left an impression that the claylands were actively avoided by Mesolithic populations.
Accounting for this seemingly fundamental difference in the distribution of Mesolithic activity is an issue that needs to be addressed through further research and fieldwork. However, it is perhaps worth noting that the majority of survey in Northamptonshire has been done using 30 m interval transects, whilst in Leicestershire the bulk of survey has used a 20 m interval. It follows that the Northamptonshire surveys will have been less likely to identify smaller sites. Prominences and well-drained locations may well have attracted repeated activity or occupations and thereby generated palimpsests of sufficient size to be well represented in all surveys. Extensive forested areas of clayland however may not have provided conditions where specific locations were reoccupied. With fewer palimpsests, the clayland site record would consist of numbers of small, spatially discrete sites. We know that in upland areas excavations have identified Mesolithic sites where the lithics were concentrated into an area of 5 m diameter or less. It may be that a narrower transect interval provides sufficient coverage to identify a good proportion of such small sites.
The impact of developer-funded work on the known Mesolithic resource in the East Midlands region has been considerable. The funding of such fieldwork outside of those areas where tradition has dictated the Mesolithic is to be found is providing data that will contribute directly to national research agendas. Programmes of fieldwork are identifying sites in previously neglected locations. It is important however that once defined, concentrated scatters should be investigated in their own right, and not just with respect to the survival of buried features. The need for typological and metrical analysis of assemblages across the region has been identified and discussed. For the region to contribute to perspectives on inter-assemblage variability, reduction sequences and raw material distributions we need to have data from well sampled lithic assemblages.
Throughout this paper reference has been made to the importance of the three relevant research priorities identified by the Prehistoric Society Working Party on the Palaeolithic and Mesolithic (1999). These essentially transitional/chronological issues can be broadly characterised as follows:
In one sense such broadly defined topics may provide a useful framework for advancing Mesolithic research. I have no doubt that these broad topics provide sufficient scope to ensure the inclusion of virtually any Mesolithic research proposal. However, in defining an East Midlands research agenda there is an opportunity to identify more specific research objectives that can be purposefully addressed. What follows is based upon the previous discussion.
As already discussed, for the generality of Mesolithic archaeology in the East Midlands to address any or all of the three transitional/chronological issues identified above will require more chronological and typological detail from the analysis of lithic assemblages. This applies equally to surface survey and excavations. It is not sufficient for scatters to be simply regarded as locations for identifying the presence or absence of features through mechanical trenching. Detailed analysis of excavated assemblages is necessary for typological, morphological and inter-assemblage research to proceed. Of course, the ever-present demands for more reliable radiocarbon dating of Mesolithic assemblages and the holy grail of finding well-preserved Mesolithic structural evidence confirms that opportunities to investigate suitable features and contexts should not be ignored. It would be desirable to see some research excavation of scatters defined by surface collection.
To achieve such objectives may require changes in the drafting of project briefs and will also require changes in fieldwork methodology. Surface collection surveys should be designed ultimately to enable various site and assemblage details to be derived from recovered lithics beyond that of simple dots on maps. A traverse and stint methodology may be efficient for the initial recognition of activity concentrations. Once identified, however, such concentrations should be plotted to a greater level of detail, with the objective of defining their extent, shape and size. Finds may or may not need to be recorded individually. The level of plotting should be appropriate for enabling the fieldwork to contribute positively to addressing the issues raised by the agenda. Account should also be taken of the impact upon rates of site recognition of having too-widely spaced traverses, given that many Mesolithic activity concentrations may be very small in diameter.
Such surveys will prove to be most useful where the fieldwork is conducted across varied geologies, geomorphologies and topographic locations to a broadly standardised level of coverage. Finally, if SMRs are to play a more constructive role in Mesolithic research, their records need to incorporate an increased level of detail on the typology and chronology of assemblages, and regarding the size and density of lithic scatters and sites.