Beruflich Dokumente
Kultur Dokumente
O south of the present estuary of the River Esk is an area of land known as the Eskmeals foreland, which has been built up by coastal processes during the Holocene (Fig. 2.1). The eastern margin of the foreland is dened by a sharp break of slope that can be traced from the River Esk north of Newbiggin to south of Stubb Place. To the landward glacial sediments underlie an undulating topography of moderate relief that rises steadily inland to over 200 m in a few kilometres. The coastal foreland itself is formed by marine, estuarine and aeolian sediments resting on an eroded surface of glacial deposits. A complex sequence of shingle ridges has been produced by the dominant northwards-directed longshore movement of littoral sediment, and in the lee of the shingle ridges estuarine sediments have accumulated. The aeolian deposits that have developed on the surface of the shingle
26
Fig. 2.1. Geomorphological map of the Eskmeals area. Key: 1. peat-lled basins; 2. present salt marsh; 3. estuarine sediments; 4. wind-blown sand sheet; 5. constructional sand dunes; 6. sharp break of slope; 7. moderate break of slope; 8. present streams; 9. former stream channels; 10. glacial sediments; 11. alluvial fans; 12. made ground; 13. shingle ridges (after Bonsall et al. 1989).
27
ridges comprise a seaward area of constructional dunes eastwards of which a large, nearly at, sand sheet has developed. A series of channels transect the glacial deposits and apparently terminate at the edge of the coastal foreland. It is likely, however, that the lower portions of some of the channels continue below the coastal sediments. These channels, particularly along their present lower portions, have sediment inlls that accumulated contemporaneously with the formation of the coastal foreland and thus preserve valuable evidence of mid- to late-Holocene environments. Traces of Mesolithic occupation occur mainly on the glacial deposits along the edge of the foreland, within a few hundred metres of the break of slope which marks the limit of the early Holocene marine transgression. Mesolithic artefacts have also been recovered from the shingle ridges surrounding the Williamsons Moss basin which are known to have formed by c.6950 cal BP. Over 30 lithic scatters with a denite or suspected Mesolithic component have been recorded to date (Fig. 2.2), exposed where the modern vegetation was disturbed by ploughing or burrowing animals. Between 1974 and 1986 sites were investigated at Monk Moors and Williamsons Moss. Development of the coastal foreland The evidence relating to sea-level change and shoreline development in the Eskmeals coastal zone has been discussed by Bonsall et al. (1986, 1989). The evolution of the coastal foreland is illustrated in Fig. 2.3 and may be summarized as follows: Early Holocene sea-level rise culminated in the contemporaneous erosion of the landward glacial deposits and formation of the innermost shingle ridge between c.7450 and 7850 cal BP. Estuarine sedimentation began in the lee of this shingle ridge (Fig. 2.3a). Mean High Water of Spring Tides was then c.2 m above that of today. Subsequently, and with only a slight fall in sea level, the greater part of the shingle ridge complex that underlies the Eskmeals foreland was constructed. The shingle ridges around Williamsons Moss were deposited prior to 6950 cal BP and a new phase of shingle ridge development which had been preceded by a net fall in relative sea level of c.1 m was underway by 4950 cal BP. This intermediate sequence of shingle ridges continued to form until after 3900 cal BP (Fig. 2.3b), their formation being accompanied by progressive abandonment of the estuarine sediments landwards of the shingle ridges. The sea subsequently fell by 0.8-1.1 m and a further phase of shingle ridge development was in progress by 3000 cal BP (Fig. 2.3c). There have been at least two periods of sand dune activity on the foreland, the rst between 3500 and 1700 cal BP, and the second sometime after c.1400 cal BP. An earlier phase of sand dune activity may be suggested by a very sandy facies near the base of the estuarine sediments closest to the main shingle ridge complex. This would have occurred at sometime between 6950 cal BP (probably after 6300 cal BP) and 3750 cal BP.
28
Fig. 2.2. Distribution of Mesolithic sites in the Eskmeals area (after Bonsall et al. 1989).
29
Settlement location and coastal change The Eskmeals sites are all situated on seaward-facing slopes overlooking stretches of former coastline that experienced predominantly estuarine conditions during the period of Mesolithic settlement, whereas settlement traces are lacking in areas where estuarine conditions were never established during the mid-Holocene for example south of Stubb Place. This pattern is repeated in coastal areas throughout Europe where late Mesolithic settlements typically are found in sheltered locations around estuaries, lagoons or marine inlets (cf. Larsson, 1991). Several authors have observed that river estuaries offer a greater abundance and variety of potential food resources than most other coastal environments (e.g. Jacobi, 1973; Clarke, 1976; Bonsall, 1981). Other resources may also have been a signicant factor in settlement location at Eskmeals. The only locally available source of int would have been the shingle deposits of the coastal foreland. These deposits would also have been the most accessible sources of sand and stones known to have been used extensively for construction purposes (see below). The main site locations at Monk Moors and Williamsons Moss would have been ideally situated to exploit these resources. Another interesting feature of the site distribution is the tendency for lithic scatters to occur along the margins of former channels cutting across the glacial sediments. This relationship is particularly evident to the landward of Williamsons Moss (Fig. 2.4) but was also noted at Monk Moors (see below). The signicance of these channels for Mesolithic settlement is uncertain. It seems doubtful that they acted as sources of fresh water. Although the channels would have carried away rainwater running off the slopes adjacent to the coastal foreland, during the mid-Holocene they are unlikely to have been occupied by perennial streams. Certain of the channel mouths, however, would have provided a means of access to the foreshore particularly in the area between Monk Moors and Stubb Place where, at the maximum of the transgression, the sea was actively eroding a cliff some 3-15 m high. It is also possible that where estuarine conditions obtained the channels had economic signicance. Salmon and sea trout are known to spawn in large numbers in the mouths of small channels and feeder streams leading into the present estuary. It has been suggested that lithic scatters often represent former rubbish dumps (Bonsall, 1992), and in some instances wet channel margins may have been chosen simply as convenient places for disposal of occupation refuse (cf. Bonsall et al., 1993). Since environmental factors such as the nearby presence of estuarine conditions appear to have had a major inuence on Mesolithic site location, it may be presumed that the chronology of settlement at the different sites will relate to coastal evolution and the progressive establishment of suitable conditions in the Eskmeals area. As estuarine conditions were rst established (between 7450 and 7850 cal BP) in the area between Monk Moors and Newbiggin, settlement may have rst occurred in that area at that time. In contrast, the Williamsons Moss site faced onto the open coast until after the construction of the shingle ridges that formed the Williamsons Moss basin at
30
STUDIES IN NORTHERN PREHISTORY
Fig. 2.3. Evolution of the Eskmeals coastal foreland. Key: 1. shingle ridges; 2. estuarine areas; 3. glacial sediments; 4. marked break of slope; 5. feature boundary; 6. streams. On each diagram the archaeological sites known to have been occupied at the relevant period are shown (after Bonsall et al., 1989, with revision).
31
c.6950 cal BP, implying that after this date that locality would have been more attractive for settlement. It is therefore interesting that the earliest dated occupation at Monk Moors was at c.7650 cal BP, while the earliest dated occupation at Williamsons Moss was at c.6350 cal BP (see below). This apparent relationship between periods of site occupation and environmental conditions suggests that as sea level fell, occupation may have continued longest in the Monk Moors-Newbiggin area where estuarine conditions lasted longest (cf. Fig. 2.3), and that Williamsons Moss would have retained its attraction for human settlement as long as an open lake occupied the area of the present moss.
32
Fig. 2.4. Lithic artefacts from Monk Moors site 1: a. microliths; b. scrapers [1-2], burin [3], bipolar (scalar) core [4] (after Bonsall, 1981).
33
materials (Bonsall, 1981). Moreover, radiocarbon dating indicates that they relate to more than one phase of occupation (Table 1). Charcoal from a hearth on Site 1 was dated as 6750 155 BP (BM-1216), indicating that the hearth relates to use of the site during the Mesolithic. This feature formed part of a linear cluster of hearths and stake holes which may belong to a single phase of occupation. Charcoal samples from a second hearth on Site 1 (isolated from the main group) and a hearth and a pit on Site 2 gave dates respectively of 3650 120 BP (BM1395), 4050 55 BP (BM-1385) and 2860 50 (BM-1386) BP, suggesting that these features relate to post-Mesolithic activity.
Table 1.
Radiocarbon dates for the Monk Moors sites. Calibrated ages have been calculated using atmospheric data from Reimer et al. (2004) and OxCal v 3.10 (Bronk Ramsey, 2005), and are shown at 95.4% probability. Lab ID BM-1216 BM-1385 BM-1395 BM-1386 Context Site 1. Charcoal from hearth. Site 2. Charcoal from hearth. Site 1. Charcoal from hearth. Site 2. Charcoal from pit.
14C
age BP
It is apparent from the radiocarbon and other evidence that the two sites investigated were occupied on a number of occasions between c.7650 and 2900 cal BP, indicating that the coastal foreland retained its attraction for settlement long after the Mesolithic period. The pattern of repeated occupation of the same site locations implied by these data also opens the possibility of more than one episode of Mesolithic occupation at Monk Moors.
34
Fig. 2.5.The Williamsons Moss area showing the excavated site in relation to local topography (after Bonsall et al., 1989).
context and showed few signs of postdepositional disturbance. It was established by means of test pits that the greatest concentration of artefacts occurred in an area roughly 50 x 30 m on the crest of the till spur. In subsequent excavation an assemblage of more than 34,000 worked pieces of int, chert and tuff, including over 600 microliths and other retouched tools, was recovered from an area of just 125 m2 (Fig. 2.5: Area A). On the crest of the spur there was no apparent build-up of alluvial material, the soils having developed directly on till. Hence the archaeological remains there were unstratied. The lithic artefacts were mixed through the upper 30-40 cm of the soil prole. This mixing was probably caused principally by bioturbation (perhaps aided by human occupation disturbance) rather than ploughing. The few structural features identied, comprising hearths, small pits and several possible stake holes (Bonsall et al., 1989: g. 40), were truncated by the modern A-horizon. In the lower-lying areas below the crest of the till spur lithic artefacts occurred mainly on the palaeo-landsurface beneath alluvial sediments, although artefact densities were very low (usually <5 per m2) and the distribution uneven. The
35
palaeo-landsurface, however, contains other evidence of human activity. Excavation exposed a number of concentrations of medium and large rounded stones forming a distinct layer on either side of the till spur. The stones contrast with those in the surrounding sediments and were probably derived from nearby shingle deposits. This layer is therefore interpreted as the remains of stone oors (stone pavement) laid down to consolidate a poorly drained clayey soil easily puddled by trampling. The main expanse of stone pavement examined lay between the crest of the till spur and the inlled channel (Fig. 2.5: Area B). From their stratigraphic context, the stone oors are presumed to relate to Mesolithic occupation (Bonsall et al., 1986, 1989). Two trenches were dug to investigate the deposits within the inlled channel (Fig. 2.5: Areas E1 & E2). Consisting largely of silty clays (see below), these deposits resemble the alluvial sediments in other parts of the site. Two types of structure were found stratied within the channel sediments and appear to relate to a palaeo-landsurface c.90 cm below that of today. One is a raft-type foundation consisting of two sets of oak branches arranged in a grille or lattice pattern, overlain by a layer of black, well-humied peat composed of twigs, branches and bark interpreted as the remains of a layer, or series of layers of birch brushwood placed on top of the timber lattice. Examples of this type of structure were found in Area E1 (Fig. 2.6: Structure 1; Fig. 2.7) and Area E2. The second type of structure consists of areas of made ground retained by a revetment of large radial sections of oak trunks. An example of this type was partially exposed in Area E1 (Fig. 2.6) and was overlain by a peaty layer containing abundant birch bark fragments interpreted as the remains of a bark oor (Fig. 2.6: Features 53 & 80). A similar feature, but lacking evidence of a birch bark oor, occurred in Area E2. It can be suggested that these structures were built to consolidate soft ground formed by the alluvial sediments inlling the channel. It seems likely that they were intended to form platforms raised above the level reached by water ponding behind the shingle ridge. Comparison with similar nds made at Mesolithic sites in southern Scandinavia can be used to suggest that the layers of birch bark associated with Structure 2 are remnants of the internal oors of one or more buildings, possibly houses, constructed on the platform and adjacent channel margin. Structures 1 and 2 are dated by six radiometric 14C determinations (Table 2). A sample of oak from the foundation of Structure 1 gave an age of 6015 75 BP (UB-2544). Two samples of birch bark from the base and upper part of the overlying layer of decayed brushwood respectively gave dates of 5555 40 BP (UB-2545) and 5650 50 BP (UB-2546). Three samples of birch bark from the two principal areas of bark ooring associated with Structure 2 gave ages of 5520 85 BP (UB-2712) and 5500 70 BP (GU-1664) [F80] and 5480 90 BP (UB2713) [F53]. The dates on bark should approximate most closely the period of construction. These dates are statistically indistinguishable and indicate that Structures 1 and 2 were built c.5550 BP. The difference in age of c.450 years indicated by the date for one of the foundation timbers of Structure 1 and those
36
Table 2.
Radiocarbon dates for the Williamsons Moss site. Calibrated ages have been calculated using atmospheric data from Reimer et al. (2004) and OxCal v 3.10 (Bronk Ramsey, 2005), and are shown at 95.4% probability. Lab ID UB-2544 UB-2546 UB-2545 UB-2712 GU-1664 UB-2713 Context Area E1. Wood (oak). Part of timber lattice of Structure 1. Area E1. Bark fragments (birch). Part of decayed brushwood covering of Structure 1. Area E1. Bark fragments (birch). Part of decayed brushwood covering of Structure 1. Area F. Bark fragments (birch). Part of bark oor on buried land surface. Area F. Bark fragments (birch). Part of bark oor on buried land surface. Area E1. Bark fragments (birch). Part of bark oor on timber and earth platform (Structure 2). Area F. Charcoal from hearth in alluvial sediments overlying buried land surface. Area B. Charcoal from hearth (F23) in alluvial sediments overlying buried land surface. Test Pit AP36. Charcoal from hearth in alluvial sediments overlying buried land surface. Area E2. Charcoal from hearth overlying Structure 5.
14C
age BP
6015 750 5650 500 5555 400 5520 850 5500 700
45004050 42503300
UB-2711 BM-1396
3756 104
25001900
UB-2568
22001920 20301610
UB-2715
for the decayed brushwood covering of this feature can be explained, in part, by the (re-)utilization of old timbers for the foundation of this structure (Bonsall et al., 1989). No diagnostic artefacts were found in direct association with either Structure 1 or Structure 2. Elsewhere on the site there are large numbers of diagnostically Mesolithic artefacts and no clearly Early Neolithic artefacts, either of stone or pottery. Moreover, the date of c.6350 cal BP assigned to the structures is signicantly older than the earliest securely dated occurrences of Neolithic artefacts in northern Britain. The structural remains in Area E1, therefore, are most likely to relate to occupation of the Williamsons Moss site by people who possessed a Late Mesolithic material culture. No dates are available for the timber structures in Area E2, but the general similarity in form, construction technique and stratigraphic context to those in Area E1 suggests that the Area E2 structures also relate to Mesolithic occupation of the site.
Fig. 2.6.Williamsons Moss, inlled channel: structural features in Area E1 after Bonsall et al., 1989).
37
38
Fig. 2.7.Williamsons Moss, inlled channel: raft-type foundation (Structure 1) in Area E1 (after Bonsall et al., 1989).
As at Monk Moors, there is some evidence for post-Mesolithic use of the Williamsons Moss site. Artefacts, hearths, and lenses of introduced beach sand and pebbles occurred within the alluvial sediments above the palaeo-landsurface, although there was no indication of major or prolonged settlement activity following the period of Mesolithic occupation. Three hearths at different levels within the alluvial sediments have given radiometric 14C ages of 3665 40 BP (UB-2568), 3755 105 BP (BM-1396) and 4925 165 BP (UB-2711) consistent with their relative stratigraphic positions (Table 2).
39
Here, c.3.50 m of organic deposits (peat and detrital mud) overlie >2.0 m of minerogenic sediments that relate to the period when Williamsons Moss was a lagoon protected by the shingle ridge which was actively forming along its seaward margin. The organic deposits between c.1.50 and 3.50 m depth were dated by a series of radiometric 14C measurements to between c.2550 and 6850 cal BP, the latter date marking the isolation of Williamsons Moss from marine inuence and the beginning of lacustrine sedimentation in the basin. The inlled channel pollen site corresponds with excavated Area E1 (Fig. 2.5, P2). The north-west wall of this trench provided a section through the channel deposits, including the organic layer resulting from the decay of Structure 1. The section was sampled for pollen analysis at a point just a few metres from the edge of the channel where the deposits were only c.1.30 m deep. The pollen prole is undated, except for the organic layer between 0.67 and 0.92 m which on 14C evidence accumulated rapidly c.6350 cal BP. The pollen record for the period from 6850 to 4500 cal BP corresponding to the nal stages of the Mesolithic and the greater part of the Neolithic suggests extensive development of mixed oak forest on the better-drained soils, with alderbirch carr in the wetter areas such as around the lake basin and on channel margins. Two phases of woodland disturbance, suggestive of clearance regeneration episodes, are recorded in the Williamsons Moss pollen prole during this period. The rst, dated between c.6500 and 5750 cal BP, is marked by uctuating percentages of oak (Quercus) pollen. Increases in the pollen of the light-demanding shrubs juniper (Juniperus) and hawthorn (Crataegus) and the appearance of ribwort plantain (Plantago lanceolata) and bracken (Pteridium sp.) suggest the creation of more open woodland in the vicinity of the lake. There is also a signicant, if temporary, decline in elm (Ulmus) pollen values in the early part of this phase beginning c.6350 cal BP. A similar fall in elm pollen values is recorded in the channel prole at about the same time, and may be the result of deliberate woodland clearance by Mesolithic people. No corresponding increase in herbaceous pollen is evident in the channel prole, which led Tipping (1994) to question whether signicant areas of open ground existed around the channel at that time (but see below). Woodland regeneration seems to have occurred by 5750 cal BP. Following a period of relative stability, a second and apparently more extensive phase of woodland disturbance is recorded in the Williamsons Moss prole between c.5650 and 4500 cal BP. The beginning of this phase is marked by a more pronounced and permanent reduction in elm pollen values, possibly corresponding to the wellknown Elm Decline which appears in pollen diagrams across north-west Europe. A review by Parker et al. (2002) concluded that the Elm Decline was due to the outbreak of disease exacerbated by man-made clearances. Signicantly, both clearance episodes recorded in the pollen proles from Williamsons Moss coincided with periods of human activity at the archaeological site implying an anthropogenic cause, although the archaeological evidence from
40
the later phase (including a hearth, some sherds of Middle Neolithic pottery, and a leaf-shaped arrowhead) appears slight compared to the extensive evidence of Late Mesolithic occupation.
41
Williamsons Moss implies that the Mesolithic population focused its activities on the coastal zone. There are, however, limits to what may be inferred from the available data. The evidence from the Williamsons Moss site suggests sedentism, but does not demonstrate it. From radiocarbon evidence alone it cannot be demonstrated that the structures within the inlled channel relate to a continuous period of occupation. The lithic scatter and areas of stone pavement cannot be dated directly. Consequently, it cannot be proved that these features are contemporaneous, nor that either relates to the phase(s) of occupation represented by the timber platforms. Similarly, it is doubtful if from pedological evidence the effects of a single large clearing for settlement could be distinguished from those of a series of small contiguous or overlapping clearings associated with separate occupation events. Tipping (1994) has suggested from palynological evidence that there is unlikely to have been a large area of open ground around the channel at the time of platform construction. This interpretation assumes that the pollen entering the channel sediments was derived largely from vegetation growing on the margins of the channel. If, however, Mesolithic settlement resulted in complete suppression of the vegetation over a substantial area around the channel, as suggested by Bonsall et al. (1989), then the pollen recruitment zone would have been considerably enlarged and the pollen entering the channel sediments at that time would reect vegetation growing some distance (possibly hundreds of metres) away. In effect, such a clearing would be invisible in the pollen record. It might, however, be reected by an increase in the occurrence of microscopic charcoal in the channel sediments. It is doubtful, therefore, whether pollen analysis by itself can provide an accurate impression of the nature and scale of vegetation clearances associated with Mesolithic settlement.
RETROSPECT
The Eskmeals Project has been an important addition to the archaeological record of north-west England. In particular, it has made a signicant contribution to our understanding of humanenvironment interactions in the Late Mesolithic. Yet two decades after the last trench was backlled, the Monk Moors and Williamsons Moss sites remain the only Mesolithic sites on the Cumbrian coastal plain to have been systematically investigated within an interdisciplinary research framework. Unsurprisingly, therefore, there are many questions concerning the Mesolithic settlement of the Cumbrian coast that remain unresolved, not least those relating to subsistence, sedentism and the transition to farming. These are an obvious target for future research.
42
ACKNOWLEDGEMENTS
This paper is dedicated to Clare Fell who took a keen interest in the Eskmeals Project. She was a regular visitor to the excavations throughout the 12-year duration of the project. I will always be grateful to Clare for her kindness and support, and for her remarkable archaeological insights.
REFERENCES
Binford, L. R., 1979, Organization and formation processes: looking at curated technologies. Journal of Anthropological Research 35(3): 255-273. Bonsall, C., 1981, The coastal factor in the Mesolithic settlement of north-west England. In B. Gramsch (ed.) Mesolithikum in Europa. Berlin: Deutscher Verlag, 451472. Bonsall, C., Sutherland, D. G., Tipping, R. M. & Cherry, J., 1986, The Eskmeals Project 1981-5: an interim report. Northern Archaeology 7(1): 3-30. Bonsall, C., Sutherland, D. G., Tipping, R. M. & Cherry, J. 1989, The Eskmeals Project: Late Mesolithic settlement and environment in north-west England. In C. Bonsall (ed.) The Mesolithic in Europe. Edinburgh: John Donald, 175-205. Bonsall, C. 1992, Archaeology of the south-west Scottish Highlands. In M. J. C. Walker, J. M. Gray & J. J. Lowe (eds) The South-West Scottish Highlands: Field Guide. Cambridge: Quaternary Research Association, 28-34. Bonsall, C., Robinson, M., Payton, R. M. & Macklin, M. G. 1993, Ln Mr, Oban. Discovery and Excavation in Scotland 1993: 76. Bronk Ramsey, C., 2005. OxCal Program v3.10. Radiocarbon Accelerator Unit, University of Oxford. Cherry, J. & Cherry, P. J., 1986, Prehistoric habitation sites in west Cumbria: Part IV, the Eskmeals area. CW2 lxxxvi: 1-17. Clarke, D. L., 1976, Mesolithic Europe: the economic basis. In G. Sieveking, I. Longworth & K. Wilson (eds) Problems in Economic and Social Archaeology. London: Duckworth, 449-481. Jacobi, R. M., 1973, Aspects of the Mesolithic Age in Great Britain. In S. K. Kozlowski (ed.) The Mesolithic in Europe. Warsaw: Warsaw University Press, 237-275. Larsson, L., 1991, Coastal adaptation in the early and middle Holocene of southern Scandinavia. Journal of the Korean Ancient Historical Society Hanguk Sanggosa Hakbo 8: 93-118. Parker, A. G., Goudie, A. S., Anderson, D. E., Robinson, M. A. & Bonsall, C., 2002, A review of the mid-Holocene elm decline in the British Isles. Progress in Physical Geography 26(1): 1-45. Reimer, P. J., Baillie, M. G. L., Bard, E., Bayliss, A., Beck, J. W., Bertrand, C. J. H., Blackwell, P. G., Buck, C. E., Burr, G., Cutler, K. B., Damon, P. E., Edwards, R. L., Fairbanks, R. G., Friedrich, M., Guilderson, T. P., Hogg, A. G., Hughen, K. A., Kromer, B., McCormac, F. G., Manning, S., Bronk Ramsey, C., Reimer, R. W., Remmele, S., Southon, J. R., Stuiver, M., Talamo, S., Taylor, F. W., van der Plicht, J. and Weyhenmeyer, C. E., 2004. INTCAL04 terrestrial radiocarbon age calibration, 0-26 kyr BP. Radiocarbon 46(3): 1029-1058. Rowley-Conwy, P., 1983, Sedentary hunters: the Erteblle example. In G. Bailey (ed.)
43
Hunter-gatherer Economy in Prehistory. Cambridge: Cambridge University Press, 111-126. Tipping, R. M., 1994, Williamsons Moss: palynological evidence for the MesolithicNeolithic transition. In J. Boardman & J. Walden (eds), The Quaternary of Cumbria: Field Guide. Oxford: Quaternary Research Association, 104-127.