Mesolithic sedentism on Oronsay: chronological evidence from adjacent islands in the southern Hebrides.

MITHEN, STEVEN

Introduction

The nature of Mesolithic subsistence and settlement on Oronsay (FIGURE 1) has been an issue of considerable debate. Mellars & Wilkinson (1987) argued that people could have been resident on that island all year round, an idea that was challenged by Mithen & Finlayson (1991) who suggested that the middens of Oronsay derived from many short-term visits by foragers principally based on the larger islands of Colonsay, Jura and Islay. Richards & Mellars (1998) presented data regarding the diet of humans represented in two of the middens, Cnoc Coig and Caisteal nan Gillean, the former of which had relied exclusively on marine protein and hence supported the notion of sedentism on Oronsay. In this contribution I will provide additional data of relevance to this issue in terms of new radiocarbon dates for Mesolithic sites on Colonsay and Islay. Calibrated values for dates cited in this text are provided at a 95.4% confidence level. These are derived from the OxCal computer programme (Ramsey 1995) employing the 1998 internationally recommended radiocarbon calibration dataset for the northern hemisphere (INTCAL98.14C: Stuiver et al. 1998).

[Figure 1 ILLUSTRATION OMITTED]

Placing Oronsay into a regional content

In the late 1980s debates about the role of the Oronsay middens within a Mesolithic settlement system were of limited value due to the rarity of well-dated Mesolithic sites on the adjacent islands to Oronsay: Jura, Colonsay and Islay. Mercer and Searight had discovered a suite of Mesolithic sites on Jura (Mercer 1968; 1970; 1971; 1972; 1974; 1980; Mercer & Searight 1986; Searight 1990; 1993). The absolute dates from these sites were either earlier than those from the Oronsay Middens, as at Lussa Wood (8195 [+ or -] 350 BP/8200-6300 cal BC, 7965 [+ or -] 200 BP/ 7500-6550 cal BC), or much later with Neolithic dates from Lussa River (4620 [+ or -] 140 BP/3650-2900 cal BC, 4200 [+ or -] 100 BP 3050-2450 cal BC) and Glenbatrick (4225 [+ or -] 230 BP/3600-2100 cal BC). The only dated Mesolithic site on Islay in the early 1990s was Newton, which was also substantially earlier than the Oronsay middens with dates of 7805 [+ or -] 90 BP/7050-6450 cal BC and 7765 [+ or -] 225 BP/7400-6100 cal BC (McCullagh 1991). With such few dates our ability to discuss the Oronsay middens in their regional context was limited. Consequently one of the rationales behind the Southern Hebrides Mesolithic Project (SHMP) was to locate and excavate further Mesolithic sites on Colonsay and Islay.

The fieldwork between 1988 and 1995 identified 22 locations on Islay and 3 on Colonsay where Mesolithic activity was implied by the nature of the chipped stone assemblages. A series of interim reports on specific aspects of the project have been published (Edwards & Mithen 1995; Finlayson & Mithen 1997; Finlayson et al. 1996; Lake et al. 1998; Mithen 1989; 1996; Mithen & Finlayson 1991; Mithen & Lake 1996; Mithen et al. 1992) and the final report of the project is in press (Mithen n.d.).

New radiocarbon dates

Thirty radiocarbon dates on wood charcoal were acquired from five Mesolithic sites: Bolsay Farm, Coulererach, Rockside and Gleann Mor on Islay, and Staosnaig on Colonsay (FIGURE 2). All but two of these (Staosnaig Q-3278, Bolsay Farm Q-3219) were derived by AMS on single entity samples. Of these dates, 14 came from Staosnaig and 10 from Bolsay Farm.

[Figure 2 ILLUSTRATION OMITTED]

The relatively large sets of dates from Staosnaig and Bolsay Farm were quite different from each other with regard to both context and range. Those from Staosnaig came from samples taken from within discrete features. All but three were pre-7000 BP--the exception being a Neolithic date (5415 [+ or -] 60 BP/4360-4040 cal BC) and two Bronze Age dates (3455 [+ or -] 60 BP/1440-1610 cal BC and 3395 [+ or -] 60 BP/1880-1520 cal BC) on charred plant material that appears to have been re-deposited into Mesolithic features. In contrast, the suite of dates from Bolsay Farm came largely from charcoal samples recovered from amidst an artefact scatter rather than from within features. These gave a range of dates from the Mesolithic (7400 [+ or -] 55 BP/6400-6090 cal BC) to the Bronze Age (3525 [ + or -] 80 BP/2040-1630 cal BC) with a concentration within the Neolithic (five dates between 4740 [+ or -] 50 BP and 4200 [+ or -] 55 BP).

The most surprising feature of the dates acquired by the SHMP is the chronological gap between 6800 and 5400 BP (FIGURE 2). As noted above, this chronological gap is also noticeable in the range of dates available from Newton and the Jura sites. This late phase of the Mesolithic is not only represented on Oronsay, but also elsewhere in Scotland such as at Morton (e.g. 6300 [+ or -] 150 BP/5550-4850 cal BC), Smittons (6260 [+ or -] 80 BP/5380-4990 cal BC) and Risga (6000 [+ or -] 90 BP/5250-4600 cal BC) (see Wickham-Jones 1994 and Finlayson & Edwards 1997 for reviews of the Scottish Mesolithic). But no dates of this period have been forthcoming from Islay, Jura and Colonsay. Indeed, the 20 dates on charcoal from the Oronsay middens, ranging from 6190 [+ or -] 80 BP/5320-4920 cal BC at Caisteal nan Gillean to 5426 [+ or -] 159 BP/4700-3800 cal BC at Cnoc Sligeach, fit almost exactly into the gap in the otherwise continuous sequence of dates on non-midden sites these islands.

A poor [sup.14]C sample or settlement discontinuity?

How should this be interpreted? The first possibility is that this absence of dates between 6800 and 5400 BP reflects no more than poor sampling, or perhaps bad luck. It is possible that the late Mesolithic foragers of this period were exploiting Islay, Jura and Colonsay as intensively as those prior to 6800 BP but were simply using as yet undiscovered or undated settlement locations. The SHMP itself discovered more than 20 localities of Mesolithic activity that remain undated and many more sites are likely to be sealed below the peat and blown sand of the island. Consequently the apparent absence of Mesolithic sites between 6800 and 5400 BP in the southern Hebrides, other than on Oronsay, may reflect no more than an inadequate sample of existing dates.

This possibility can be partly evaluated by using the palynological and charcoal records from sediments as these provide an additional monitor of human activity. Several cores have been extracted from Loch a'Bhogaidh located about 600 m from Bolsay Farm (Agnew et al. 1988; Edwards & Berridge 1994). These show possible human impacts on the vegetation for the periods 7500-7300 BP and c. 7000-6500 BP, and then between 5230 BP and 3610 BP. Such periods of vegetation disturbance fit extremely well with the Mesolithic radiocarbon dates from Bolsay Farm -- 7400 [+ or -] 65, 7250 [+ or -] 145, 6810 [+ or -] 55 BP -- and the sequence of later prehistoric dates between 4740 [+ or -] 50 BP -- 3525 [+ or -] 80 BP. Just as the period 6800-5400 BP is missing from the Bolsay Farm radiocarbon dates, so too is there no evidence for vegetation disturbance at that time.

There is evidence, therefore, for a distinctive final phase of the Mesolithic in the southern Hebrides during which the majority, perhaps even the whole, of Mesolithic settlement appears to have occurred on the tiny island of Oronsay.

Why might this have occurred? Why would Mesolithic people have abandoned the larger islands of Islay, Jura and Colonsay, which had been exploited for as much as 2000 years, for the tiny, exposed island of Oronsay? There is no indication from the palaeoenvironmental data for any resource decline on the larger islands at c. 6800 BP. According to the Loch a'Bhogaidh core, and others analysed by the Southern Hebrides Mesolithic Project, this was a period when mixed oak woodland was established across the landscape and when alder was beginning to become widespread. A study of sea-level change in the southern Hebrides has indicated that the marine transgression had commenced at 9000 BP and created a tidal strait separating the Rinns from the rest of Islay for 7000 years (Dawson et al. 1998), but there appears no substantial environmental change at c. 6800 BP that may have initiated a shift in settlement pattern.

There can be no question that the resource base of Oronsay between 6800 and 5400 BP was markedly less productive than that of the larger adjacent islands: the island would have lacked the diverse range of terrestrial animals and plants available in the woodlands of Islay and Jura that appear to have been very attractive to Mesolithic people before 6800 BP and those of the Neolithic after 5400 BP. The narrow range of resources available on Oronsay appears to be reflected in the total reliance on marine protein in the diet of the Cnoc Coig individual(s). Consequently it is difficult to find any ecological rationale for Mesolithic people between 6800-5400 BP to have become sedentary on Oronsay.

Discussion

There are now three lines of evidence that can be used to support the idea of sedentary settlement on Oronsay:

1 Faunal evidence from saithe otoliths suggesting that all seasons of the year are represented on the island (Mellars & Wilkinson 1980);

2 Stable isotope evidence from Cnoc Coig human bones indicating a total reliance on marine protein (Richard & Mellars 1998);

3 The absence of any dated Mesolithic site on Colonsay, Jura and Islay within the time frame of the Oronsay middens, 6800-5400 BP.

There are two lines of evidence which can challenge the interpretation of sedentism on Oronsay: the small size and restricted material culture of the middens, and the lack of an ecological rationale for Mesolithic foragers to restrict their settlement to a tiny island -- an island with a markedly restricted resource base when compared to that of the adjacent large islands.

At present there are three possible scenarios to account for these conflicting sets of evidence:

1 Between 6800 and 5400 BP Mesolithic people abandoned their exploitation of Islay, Jura and Colonsay and became sedentary on Oronsay. Their motivation for this appears to have been more ideological than ecological. In this scenario the Caisteal-nan-Gillean II human bone, which Richards & Mellars (1998) argue is indicative of a diet combining protein from terrestrial herbivores and marine mammals, would need to relate to a period either at the start or at the end of this period of sedentism, whereas the Cnoc Coig human bones would date from a period when sedentism on Oronsay was fully established.

2 A second alternative would have Late Mesolithic foragers continuing a highly mobile lifestyle but shifting their exploitation to other regions, such as Mull and the Scottish mainland. Oronsay became (or more probably continued as) a locality for specialized activities which were undertaken at various times of the year with the middens accumulating from a large number of intermittent visits to the island.

3 A third possibility is that between 6800 and 5400 BP the Mesolithic settlement pattern continued with little change to that which had existed since the first occupation of the southern Hebrides, currently dated to c. 8000 BP with the lack of dated sites on Islay, Jura and Colonsay for the period 6800-5500 BP arising from an unrepresentative sample of existing radiocarbon dates.

Although the southern Hebrides now has a regional Mesolithic data-base that is unparalleled in Britain it remains difficult to select which of these three scenarios is the most likely. My preference is for the second. But future fieldwork on the islands should be able to differentiate between them, as they have clear implications for what type of sites should be found by future fieldwork in the southern Hebrides and adjacent regions. Whatever the nature of the settlement pattern in the period 6800-5400 BP it is evident from the radiocarbon dates acquired by the SHMP, notably those from Bolsay Farm, that the early Neolithic people of this region re-used locations which had been favoured by Mesolithic people prior to 6800 BP. In this regard a late Mesolithic phase of predominantly marine exploitation as represented on Oronsay may have been a short-term anomaly from an otherwise mixed terrestrial and marine based economy that lasted throughout the Mesolithic and Neolithic.

Acknowledgements. I am grateful to Bill Finlayson for discussion of the issues raised in this paper, to Sturt Manning for help with the calibration of radiocarbon dates and to Petra Dark for comments on a previous version of this manuscript. The views expressed in this paper are mine alone, and not necessarily shared by my colleagues on the Southern Hebrides Mesolithic Project. I am also grateful to Historic Scotland for their support of the SHMP, especially with regard to secruing radiocarbon dates, and to Patrick Ashmore for discussion as to their interpretation.

References

AGNEW, A.D.Q., K.J. EDWARDS, K.R. HIRONS, P.D. HULME & M.C.F. PROCTOR. 1988. The mire of Loch a'Bhogaidh, Islay, Transactions of the Botanical Society of Edinburgh 45: 187-201.

DAWSON, S., A.G. DAWSON & K.J. EDWARDS. 1998. Rapid Holocene relative sea level changes in Gruinart, Isle of Islay, Scottish Inner Hebrides, The Holocene 8: 183-95.

EDWARDS, K.J. & J.M.A. BERRIDGE. 1994. The Late-Quaternary vegetational history of Loch a'Bhogaidh, Rhinns of Islay SSSI, Scotland, New Phytologist 128: 749-69.

EDWARDS, K.J. & S.J. MITHEN. 1995. The colonization of the Hebridean Islands of Western Scotland: evidence from the palynological and archaeological records, World Archaeology 26(3): 348-61.

FINLAYSON, B. & K.J. EDWARDS. 1997. The Mesolithic, in K.J. Edwards & I.B.M. Ralston (ed.), Scotland: environment and archaeology, 8000 BC-AD 1000: 109-25. Chichester: John Wiley & Sons.

FINLAYSON, B., N. FINLAY & S.J. MITHEN. 1996. Mesolithic chipped stone assemblages: descriptive and analytical procedures used by the Southen Hebrides Mesolithic Project, in Pollard & Morrison (ed.): 252-66.

FINLAYSON, B. & S.J. MITHEN. 1997. The microwear and morphology of microliths from Gleann Mar, in H. Knecht (ed.), Projectile technology: 107-29. New York (NY): Plenum Press.

LAKE, M.W., P.E. WOODMAN & S.J. MITHEN. 1998. Tailoring GIS for archaeological applications: an example concerning viewshed analysis, Journal of Archaeological Science 25: 27-38.

MCCULLAGH, R. 1991. Excavation at Newton, Islay, Glasgow Archaeological Journal 15: 23-50.

MELLARS, P.A. 1987. Excavations on Oronsay: Prehistoric human ecology on a small island. Edinburgh: Edinburgh University Press.

MELLARS, P.A. & M.R. WILKINSON. 1980. Fish otoliths as indicators of seasonality in prehistoric shell middens: the evidence from Oronsay (Inner Hebrides), Proceedings of the Prehistoric Society 46: 19-44,

MERCER. J. 1968. Stone tools from a washing-limit deposit of the highest Post-Glacial transgression, Lealt Bay, Isle of Jura, Proceedings of the Society of Antiquaries of Scotland 100: 1-46.

1970. Flint tools from the present tidal zone, Lussa Bay, Isle of Jura, Argyll, Proceedings of the Society of Antiquaries of Scotland 102: 1-30.

1971. A regression time stone workers camp 33 ft OD Lussa River, Isle of Jura, Proceedings of the Society of Antiquaries of Scotland 103: 1-32.

1972. Microlithic and Bronze Age camps 75-26 ft OD N. Carn. Isle of Jura, Proceedings of the Society of Antiquaries of Scotland 104: 1-22.

1974. Glenbatrick waterhole, a microlithic site on the Isle of Jura, Proceedings of the Society of Antiquaries of Scotland 105: 9-32.

1980. Lussa Wood I: the late glacial and early postglacial occupation of Jura, Proceedings of the Society of Antiquaries of Scotland 110: 1-31.

MERCER, J. & S. SEARIGHT. 1986. Glengarrisdale: confirmation of Jura's third microlithic phase, Proceedings of the Society of Antiquaries Of Scotland 116: 41-55.

MITHEN, S.J. 1989a. New evidence for Mesolithic settlement on Colonsay, Proceedings of the Society of Antiquaries of Scotland 119: 33-41.

1996. Mesolithic settlement and raw material avaliability in the southern Hebrides. in A. Fischer (ed.), Man and sea in the Mesolithic: coastal settlement above and below present sea level: 265-73. Proceedings of the International Symposium Kalundborg, Denmark 1993. Oxford: Oxbow. Monograph 53.

(Ed.). N.(t. Hunter-gatherer landscape archaeology: The Southern Hebrides Mesolithic Project 1988-1998. Cambridge: McDonald Institute.

MITHEN, S.J. & B. FINLAYSON. 1991. Red deer hunters on Colonsay? The implications of Staosnaig for the interpretation of the Oronsay middens. Proceeding of the Prehistoric Society 57: 1-8.

MITHEN, S.J., B. FINLAYSON, N. FINLAY & M. LAKE. 1992. Excavations at Bolsay Farm, a Mesolithic site on Islay, Cambridge Archaeological Journal 2: 242-53.

MITHEN, S.J. & M.W. LAKE. 1996. The Southern Hebrides Project: reconstruction of Mesolithic settlement in western Europe, in Pollard & Morrison (ed.): 123-51.

POLLARD, A. & A. MORRISON (ed.). 1996. The early prehistory of Scotland. Edinburgh: Edinburgh University Press.

RAMSEY, C.B. 1995. Radiocarbon calibration and analysis of stratigraphy: The OxCal program, Radiocarbon 37: 425-30.

RICHARDS, C. & P.A. MELLARS. 1998. Stable isotopes and the seasonality of the Oronsay middens, Antiquity 72:178-84.

SEARIGHT, S. 1990. Mesolithic activity at Cam southern raised beach, Isle of Jura, Proceedings of the Society of Antiquaries of Scotland 120: 7-16.

1993. Lussa Bay, Isle of Jura, Argyll: a note on additional tools, Proceedings of the Society of Antiquaries of Scotland 123: 1-8.

STUIVER, M., P.J. REIMER, E. BARD, J.W. BECK, G.S. BURR, K.A. HUGHEN, B. KROMER, E.G. MCCORMAC, J. VAN DER PLICHT & M. SPURK. 1998. INTCAL98 radiocarbon age calibration, 24,000-0 cal BP, Radiocarbon 40: 1041-83.

WICKHAM-JONES, C.R. 1994. Scotland's first settlers. London: Batsford Press.

STEVEN MITHEN(*)

(*) Department of Archaeology, University of Reading, Whiteknights, Reading RG6 6AA, England. s.j.mithen@reading.ac.uk Received 9 June 1999, accepted 23 September 1999, revised 15 November 1999.