Objects for an Island World, North Roe Felsite, Shetland

Home , Knap of Howar, Lambay Island, Langdale axe industry, North Roe, Trefignath

Objects for an island world, North Roe Felsite, Shetland

Cooney, G., Ballin, T., & Megarry, W. (2012). Objects for an island world, North Roe Felsite, Shetland. In D. Mahler (Ed.), The Border of Farming and the Cultural Markers (pp. 87-93)

Published in: The Border of Farming and the Cultural Markers

Queen's University Belfast - Research Portal: Link to publication record in Queen's University Belfast Research Portal

General rights Copyright for the publications made accessible via the Queen's University Belfast Research Portal is retained by the author(s) and / or other copyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associated with these rights.

Take down policy The Research Portal is Queen's institutional repository that provides access to Queen's research output. Every effort has been made to ensure that content in the Research Portal does not infringe any person's rights, or applicable UK laws. If you discover content in the Research Portal that you believe breaches copyright or violates any law, please contact [email protected].

Download date:26. Sep. 2021 The Border of Farming and the Cultural Markers

Short papers from the network meeting in Lerwick, Shetland September 5th – 9th 2011

Edited by Ditlev L. Mahler Northern Worlds The National Museum of Denmark The Border of Farming and the Cultural Markers

Short papers from the network meeting in Lerwick, Shetland September 5th - 9th 2011

Edited by Ditlev L. Mahler.

ISBN: 978-87-7602-193-1

Funded by the Augustinus Foundation

You may freely copy articles from this publication but the author must be acknowledged, no changes must be made and you may not sell any part of this work.

The digital version of this publication is at: http:// nordligeverdener.natmus.dk/forskningsinitiativer/ samlet_projektoversigt/shetlandsoeerne_land- brug_paa_graensen_4000_3000_fvt/

Front cover: Europe’s northern most still existing passage grave at Caldback, Unst, seen from the East. Photo by Ditlev L. Mahler, summer 2011 The Border of Farming and the Cultural Markers

Short papers from the network meeting in Lerwick, Shetland September 5th – 9th 2011

Edited by Ditlev L. Mahler

Northern Worlds The National Museum of Denmark Copenhagen 2012 2

Ortholit at Pinhoulland. D. Mahler photo Contents

• Foreword 4 H. C. Gulløv

• Neolithic Shetland: a view from the “mainland” 6 Alison Sheridan

• Pinhoulland – a multi period site from West Mainland, 37 Shetland | Some Features of the Neolithic of Shetland Ditlev L. Mahler

• Sacred work: cultivating the soil in prehistoric Shetland 53 Jenny Murray

• The distribution of worked felsites 62 – within and outwith Neolithic Shetland Torben B. Ballin

• Axes from islands: the role of stone axeheads from 79 insular sources in the Neolithic of Ireland and Britain Gabriel Cooney, Stephen Mandal, Emmet O’Keefe, Graeme Warren, Torben Ballin and Will Megarry

• Multi-period constructions of megalithic tombs 100 – and the megalithic tombs in Shetland Flemming Kaul

• The Northermost Bronze Age farms 122 Preben Rønne

• Neolithic Shetland: Peopling an empty area 4000-3000 BC? 132 Ditlev L. Mahler

• Contributors 144

Nordlige Verdener Shetlandsprojekt 4

Northern Worlds The interdisciplinary research initiative of the National Museum

Hans Christian Gulløv Coordinator, senior researcher

Ethnographic Collection, National Museum of Denmark

The National Museum of Denmark has National Museum. The projects are orga initiated its most comprehensive inter- nized within three main research areas disciplinary research venture so far: defined to create sufficiently broad, Northern Worlds. Between 2009 and dynamic and interdisciplinary research 2013, the programme will produce and environments for the topics. The project communicate new knowledge on the re- Shetland – the Border of Farming 4000- lationship between people and environ- 3000 BC is part of the research area: ment over the last 15.000 years in ways relevant to the present, with its notable Farming on the edge: climatic changes. Cultural landscapes of the North The expansion of agriculture into the The research initiative Northern Worlds temperate and sub-arctic zones of the combines and coordinates the expertise planet represents a more than 6.000 year of the National Museum within the dis- long narrative, characterized by repeated ciplines of archaeology, history, ethno advances followed by stagnation. Farm- graphy, conservation and natural science ing on the edge focuses on periods and (environmental history). areas with large potential for the creation of new knowledge on agricultural Northern Worlds has 20 different sub- advances and their associated social projects, which are led by researchers structures and ideologies. The ultimate from the various research units at the boundaries of farming communities in different parts of Scandinavia and the tion is the main funder of Northern North Atlantic are explored. The project Worlds. It is with pleasure that it is pos- Shetland – the Border of Farming 4000- sible to present the second report from 3000 BC is part of this initiative. the network conference held in Lerwick, Shetland, September 5th-9th 2011 within In economical terms, Northern Worlds is the scientific project Shetland – the Border the National Museum’s greatest research of Farming 4000-3000 BC. initiative ever. The Augustinus Founda-

Nordlige Verdener Shetlandsprojekt 6

Neolithic Shetland: a view from the ’mainland’

Alison Sheridan

Introduction 1984), there had been nothing to com- In comparison with the rest of Scotland, pare with the flurry of surveying and ex- where our understanding of the Neolith- cavation that took place during the late ic period has advanced considerably as 1940s, 1950s and 1960s, with the inten- a result of excavation and targeted re- tion of clarifying Neolithic settlement search, Neolithic Shetland has remained and funerary practices (e.g. Calder 1950; something of an enigma. This is largely 1956; 1963; Henshall 1963). because the focus for excavation and research in Shetland over the last quar- However, there have been signs of a re- ter century has tended to be on later pe- cent growth of interest in Shetland’s Ne- riods, principally the Iron Age and Viking olithic, not least because of discoveries and Norse period, and because commer- at West Voe, Sumburgh in 2002, 2004 cial and rescue archaeology has not op- and 2005 when Mesolithic and Neolithic erated on the same scale as on the Scot- middens, exposed through coastal ero- tish ‘mainland’, leading to fewer sion, were investigated (Melton 2005; discoveries. Until recently, very little Melton & Nicholson 2004 and see be- Neolithic-orientated excavation or re- low). That small-scale excavation spark search had been carried out since Alas- ed Bradford University’s research project dair Whittle’s excavations at the Scord on the Mesolithic-Neolithic transition in of Brouster settlement between 1977 Shetland, which has included palaeo and 1979 (Whittle et al. 1986) and Roy environmental investigation near West Ritchie’s investigation of felsite sources Voe and an osteological and isotopic and their products (R. Ritchie 1968; analysis of the human remains from the 1992). With the exception of Gordon cist found at Sumburgh Airport in 1977 Parry’s West Burra survey (Hedges (Edwards et al. 2009; Gilmore & Melton 7

2011; Melton 2008; 2009; Melton & If this controversial development were Montgomery 2009). Renewed interest in to go ahead, this strategy would include the exploitation of felsite for the manu- the excavation of several cairns and set- facture of axeheads and knives (Ballin tlements of presumed Neolithic date. 2011a; 2011b) has led to the current re- The third initiative is, of course, the Na- search project as outlined by Ballin and tionalmuseet’s Farming on the Edge: Cooney elsewhere in this volume, while Cultural Landscapes of the North Project the excavation of sites of probable Neo- (Mahler & Andersen 2011), which has lithic (and later) date on the Hill of Crook- given rise to this volume and has acted setter near Sullom Voe in 2010 and 2011, as a catalyst for fresh research, includ- in advance of the construc tion of a gas ing detailed field survey of several set- processing plant by Total E&P UK, prom- tlements and related land divisions by ises to produce important new evidence Ditlev Mahler. This work complements regarding Neolithic settlement in that part the survey work undertaken by Val Turn- of Shetland (Ballin 2011c; Brend 2010; er (Turner 2011) and the aerial photo- Brend & Barton 2011). Furthermore, in graphic work undertaken by David Cow- the last three years there have been ley of the Royal Commission on the three initiatives that focus specifically Ancient and Historical Monuments of on Shetland’s Neolithic: the first is part Scotland in 2010 at Sandwick Bay and of the Scotland-wide ScARF (Scottish the Walls area (RCAHMS 2010: 194); Archaeological Research Framework) the results of all of this survey work Project, undertaken for Historic Scotland need to be followed by excavation in or- by the Society of Antiquaries of Scotland der to determine how many of the sites (http://www.socantscot.org/scarf.asp), are of Neolithic date. All these develop- and involves a critical appraisal of our ments mean that the time is ripe to as- current state of knowledge and out- sess what we can say, and what we standing research questions. The second need to find out, about this fascinating is the development of The Neolithic but under-researched part of Shetland’s Heart of Shetland Heritage Strategy, un- past. This contribution constitutes an in- dertaken by AOC Archaeology for Viking itial attempt to do just that. Energy in 2010 and relating to plans for a major wind farm development around Beginnings: the Earliest Neolithic Voe in Central Mainland (http://www. and prior human activity shetland.gov.uk/planningcontrol/docu- While it has been claimed, on palyno- ments/AppendixA13.6TheNeolithicHear logical grounds, that there was human tofShetlandHeritageStrategy.pdf). presence in Shetland as early as c. 6000

Nordlige Verdener Shetlandsprojekt 8

BC (Edwards et al. 2009: 117), the earli- lacks a clear cutting edge…Neverthe- est unequivocal evidence comes from less, the consensus among those who the aforementioned West Voe site (see have studied this implement is that it is a above for references) and consists of a Mesolithic type, and the question of its shell midden, comprising mainly oyster provenance is critical. This was a surface shells along with bones of seals and find “embedded with other pebbles in a birds, that lay at the bottom of the se- patch of bare ground from which the peat quence of deposits. Shell and charcoal had been eroded” on “the summit of a from this lowest, Mesolithic, midden knoll…about 800 yards west-north-west produced radiocarbon dates (with the of the naval huts at North Haven” (Cum- shell dates calibrated with regard to the ming 1946). One would have to specu- marine offset) of between 4200 and late quite hard to produce circumstances 3600 BC (see Melton 2009 for details). in which this find could be a modern im- The fact that this site was discovered as port, and yet there is absolutely no other a result of coastal erosion reminds us evidence for human activity in Fair Isle that the record for early human activity before the Neolithic…For the moment in Shetland is likely to have been trun- the conclusion must continue to be that cated, not only through erosion but be- “it would seem rash to suggest this as cause of a 9 metre sea level rise since c. an indigenous Mesolithic find” (Saville 5500 BP (as measured near Lerwick, and 1994)’ (Saville 2000, 94–5). as attested by submerged logs of 6th millennium BC date found in Lerwick har- The evidence relating to the earliest Ne- bour: Melton 2009: 185; 2010). olithic presence in Shetland is similarly sparse and, in some cases, problematic. Other evidence for Mesolithic activity is Indeed, defining what is meant by ‘Neo- more equivocal, as Edwards et al.’s 2009 lithic’ in any particular area can be discussion of the palynological record fraught with difficulties since we know, makes clear; and, as for the oft-cited from elsewhere in Europe (e.g. southern Mesolithic flint core axehead from Fair Scandinavia), that there can be selec- Isle (Cumming 1946), there remains un- tive adoption of traits associated with a certainty about its status (e.g. Saville farming way of life – such as the use of 2000). Since so much has been written pottery – without switching subsistence about this object, it would be prudent to strategies from hunting/gathering/fish- cite Saville’s comments: ‘…the Fair Isle ing. Be that as it may, the earliest evi- implement is not as typologically explicit dence in Shetland for any trait normally as one would wish, since in this case it associated with a Neolithic way of life 9

consists of just a handful of small, fea- episode of burning, not necessarily an- tureless sherds (see below) which were thropogenic in nature. found, along with bones from ‘a large terrestrial ungulate’ (species unspeci- However, what can be said, from taking fied) and limpet and mussel shells, lying a broader perspective of late 5th and ear- at the top of the aforementioned oyster ly 4th millennium developments, is that midden at West Voe. A radiocarbon date Shetland does not seem to have formed from the mammal bone (OxA-14242) in- part of either of the two initial strands dicated that these were deposited c. of Neolithisation that have been identi- 3700–3600 cal BC1. (See Melton 2009 fied on the Scottish ‘mainland’. As this 1 Note: with regarding the dating of this site.) author has argued elsewhere (e.g. the exception of the West Sheridan 2010), these appear to have in- Voe dates The other evidence that has been cited volved small-scale migration of farming (where Nigel to support the idea of a Neolithic pres- groups from northern France, followed Melton’s original marine ef- ence in Shetland prior to c. 3600 BC is by acculturation of the indigenous hunt- fect-corrected more problematic. Firstly, while Edwards er-gatherer-fisher communities. The first calibrations et al. (2009) argued that a long phase of strand featured movement up the Atlan- are cited), all radiocarbon forest reduction from c. 3910 BC at Loch tic façade from the Morbihan area of dates cited of Gards could relate to agricultural ac- Brittany at some time between c. 4300 here have been tivities, no cereal-type pollen was found and 4000 BC, and is most clearly attest- calibrated using OxCal 4.1 in the sampled material and this inter- ed at Achnacreebeag, Argyll & Bute, and are cited pretation can only be provisional. Sec- where Breton-style pottery was found in at their 95.4% ondly, a single date of 5050±85 BP a Breton-style megalithic monument probability value, rounded (4030–3660 cal BC, CAR-253), obtained (featuring a closed chamber succeeded to the nearest from a mixed sample of charcoal (de- by a simple passage tomb, the pottery decade. scribed as ‘birch, etc.’: Whittle et al. having been found in the latter: J.N.G. 1986: 37, table 1) from House 2 at Scord Ritchie 1970). of Brouster, poses problems. The sample was believed to relate to the initial This ‘strand’ of Neolithisation gave rise use of the stone house, and yet it is sig- to the tradition of building passage nificantly earlier than the four dates re- tombs in Scotland and Ireland. The sec- lating to the pre-house occupation and ond strand – the so-called ‘Carinated to the construction of the house (ibid.). Bowl [or ‘CB’] Neolithic’ – is most likely No explanation for this anomaly was of- to have originated in or around the Nord- fered and it may well be that this repre- Pas de Calais region in the far north of sents residual material from an earlier France, and will have arrived as far north

Nordlige Verdener Shetlandsprojekt 10

as Caithness at some time between the earliest form of Carinated Bowl pot- 4000 and 3800 BC. (See Sheridan 2007a tery, although the latter does encom- and 2010 for a description of its key pass some vessels of comparable thick- characteristics, and Sheridan 2012 for a ness and coarseness. review of the recent Bayesian modelling of the available dates for the initial CB There is some evidence to suggest that Neolithic by Whittle et al. 2011: chapter the Neolithisation of Shetland could 14 and fig. 14.177.) Despite their small have occurred as part of a secondary ex- size and lack of diagnostic characteris- pansion of the farming lifestyle and its tics, the handful of sherds from West associated traditions from western Voe cannot comfortably be assigned to Scotland, perhaps during the late 38th or either the Breton ceramic tradition or to 37th century BC. This expansion is at- 11

tested by the spread of passage tomb ward in support of this hypothesis is the building to the Western Isles, the north- discovery, somewhere in Shetland, of ern mainland of Scotland and the North- two axeheads of porcellanite from northern Isles, as reflected in the distribution east Ireland (Clough & Cummins 1988: of what Audrey Henshall has described 240, SHE 3 and R. Ritchie 1992: 216). It as ‘Orkney-Cromarty’ (or ‘Orkney-Cro- is known that axeheads of this material marty-Hebrides’) passage tombs (Hen- travelled northwards along the Atlantic shall 1963: 304–57 and 495–534; cf. façade of Scotland – as attested, for in- Davidson & Henshall 1991: fig. 2 and stance, by a complete axe found at Shul- 80–85). As suggested above, its origins ishader in Lewis (Sheridan 1992), its ro- lie in the tradition of passage tomb saceous wood haft radiocarbon dated to building as established by the Breton, 4470±95 BP (3490–2910 cal BC; cf. Atlantic façade strand of Neolithisation. Sheridan 1986). The arrival of these axe The evidence that this expansion heads in the hands of immigrant farm- reached as far as Shetland consists prin- ers from western Scotland is a possibil- cipally of the passage tombs of relative- ity; the sources of porcellanite are ly simple design on Shetland, and in par- known to have been exploited during ticular the examples on Ronas Hill and the first half of the 4th millennium. Swart-Houll with their simple polygonal chambers, which show some points of Furthermore, if this hypothesis regard- similarity of design with, for example, ing the secondary expansion of Neolith- Tulach an t’Sionnaich in Caithness (Fig. 1). ic practices and traditions is correct, It is unfortunate that no dating evidence then one can add that there is no evi- is available from the Shetland tombs, dence to suggest that it arrived in Shet- and one priority for future fieldwork land via Caithness or Orkney; instead, a could be their investigation. The other direct arrival from western Scotland is a piece of evidence that could be put for- distinct possibility. The reason for this claim – which is an important proposi- tion since it informs our understanding Fig. 1: Possibly the earliest megalithic monuments of the genesis and subsequent develop- in Shetland, and comparanda. Top left: Ronas Hill; ment of the Shetland Neolithic – is that middle, Swart-Houll; right, Round Hill; bottom left, Balvraid, Highland (Inverness-shire); middle, there are no signs of the stalled cham- Tulach an t’Sionnaich (first phase structure), High- ber design that forms part of the earliest land (Caithness); right, Marrog h, North Uist. megalithic tradition in both Caithness Plans from Henshall 1963; Henshall 1972; and Davidson & Henshall 1991, reproduced with per- and Orkney; similarly, there is no trace mission of Edinburgh University Press. of the subsequent (probably c. 3500 BC)

Nordlige Verdener Shetlandsprojekt 12

process of competitive aggrandisement was constructed immediately above in monument building, as seen in the that pit. Further north, on the Hill of massive horned cairns of both these re- Shurton near Lerwick, organic material gions (e.g. at Camster Long, Caithness: – presumably charcoal – preserved in Davidson & Henshall 1991: plate 21; cf. sediment immediately underlying a long, Davidson & Henshall 1989: fig. 31 for sub-peat stone wall produced a radiocar their distribution). bon date of 4740±50 BP (3640–3370 cal BC, UB-2122: Whittington 1978. Note that Developments in Shetland c. 3600- the precise nature of the dated sample 3300 BC: Early to Middle Neolithic is not specified in the publication). Whit- In the same way that much remains to tington argued that the concentration of be discovered about the initial Neolithi charcoal at this depth could have related sation of Shetland, so the narrative for to moor-burning to maintain pasture, and succeeding periods is patchy, and certain this is indeed a possibility; he also noted misconceptions need to be dispelled. that the pollen record revealed that the landscape around the time when the Positive evidence for activity between c. wall was built was a heathland, virtually 3600 and c. 3300 BC has been obtained free of trees. Another palynological study at West Voe and on Shurton Hill; addi- undertaken during the 1970s, at Mur- tional evidence, from Modesty, will be raster, claimed a mid-4th millennium date discussed below. At West Voe, the for the establishment of a continuous aforementioned episode of activity fea- record for ribwort plantain (Plantago turing the deposition of pottery and as- lanceolata), indicating an open environ- sociated bone and limpet shells seems ment and possibly farming activity to have been followed by an episode of (Jóhansen 1976; 1985). sand deposition, probably representing a storm c. 3600–3500 BC (Melton 2009, A new and very important piece of evi- 188). Thereafter, a cow tooth and some dence for activity that may belong with- cockle shells were deposited, one of the in the third quarter of the 4th millennium latter being radiocarbon dated (allowing BC has just been obtained, for the Farm- for the marine effect) to c. 3500–3300 ing on the Edge Project. This is the radio cal BC (OxA-14161: Melton 2009: 188). carbon date of 4580±35 BP (3500–3110 Possibly around the same time, a pit was cal BC, SUERC-37997), obtained from dug, with a cockle shell from that pit charcoal of short-lived Maloideae species producing a date of c. 3500–3100 cal BC associated with a deposit of nine felsite (OxA-14180: ibid.); and then a stone wall axeheads and 13 polished Shetland 13

knives, found in a long ‘knoll’ – probably on some other Shetland knives (Mahler a humanly-made mound, c. 18 x 27 me- 2010); whether it represents a re-sharp- tres in size – at Modesty, near Bridge of ening of a worn cutting edge, or a delib- Walls (Figs. 2–5; Kinghorn 1895). The erate act of ‘decommissioning’, is a material was acquired by the (then- question that requires further investiga- named) National Museum of Antiquities tion, by experimental archaeology and of Scotland in 1894. This is a particularly microwear analysis. Among the axe- interesting assemblage for several rea- heads (Fig. 3) are two (marked AC 494 sons – not least because it is the only and AC 500) which may well be re-used well-dated instance of axeheads and fragments of larger axeheads, suggest- knives found together. The knives are in ing that they were not new when depos- varying degrees of completeness, and ited; another (AC 495) may have been not all are fully polished. Eight out of the deliberately broken, and possibly slightly 13 knives (Fig. 2) show clear sign of re- burnt, in antiquity. As for the rest, the touch along one edge, with no obvious largest (c. 240 mm long, AC 493) is also evidence of subsequent wear (such as the finest, and another (AC 508, c. 195 might be expected had they been reused mm long) is relatively fine; but most of as scrapers). This feature has been noted the others could fall within the category

Fig. 2: Shetland knives of felsite from Modesty. Photo: National Museums Scotland.

Nordlige Verdener Shetlandsprojekt 14

Fig. 3: Axeheads of felsite from Modesty. Photo: National Museums Scotland.

of ‘workaday’ axeheads, and several of a rubbing stone found nearby fitted the these retain natural irregularities in their hollow in the quern’s upper surface. surface. (These stone objects, which were found ‘embedded in the charred wood’ – the Remains of three large, coarse pots of same level as the axeheads, and pre- steatitic clay, ranging in estimated rim sumably also the knives – were unfortu- diameter between c. 270 mm and c. 370 nately not acquired by the Museum.) mm, were also found in the ‘knoll’ (Fig. Sherds of the other two pots were also 4). One of these was reportedly found found in the ‘knoll’, but their spatial rela around 18 inches (c. 46 cm) from one of tionship to the axeheads and knives was the axeheads (Fig. 3, Reg. No. NMS not recorded. All three vessels contain X.AC 508), crushed flat, and partly cov- appreciable amounts of burnt-out grass, ered by what sounds to be a stone quern; which will have been used as a filler; in 15

Fig. 4: Potsherds from Modesty. Photo: National Museums Scotland.

Fig. 5: Lumps of burnt potter’s clay (right) and fragments of carbonised roundwood (left) from Modesty. Photo: National Museums Scotland.

Nordlige Verdener Shetlandsprojekt 16

other words, to employ a commonly- sets this find apart from the hoard finds used term, they had been grass-temper- such as the Shurton Brae knife hoard, ed. They have squared-off rims and ap- and the presence of cooking pots, pot- pear to have been uncarinated, although ter’s clay and a quern and rubber all there are hints of a slight carination on point towards possible domestic activi- one sherd. Encrusted organic residue on ty. Clearly there had been a considera- the interior of the lower part of one pot, ble amount of burning at this site, which together with burning of its exterior, led to the formation of a 4-5 inch (c. 10– shows that it had been used for cooking; 13 cm)-thick layer of charred wood rest- and the presence of repair holes in sev- ing on the subsoil, and a slightly thicker eral sherds shows that at least two of layer of ‘yellow peat ashes’ above this. these pots had been used for some time It would be worth re-investigating the when deposited. Also present was one mound, if any of it remains; the house- large and one or two smaller lumps of holder had disturbed it in making a gar- burnt potter’s clay (Fig. 5.1), kneaded den for his house. into dough-shaped pieces (and retaining finger impressions). These did not con- The relatively broad spread of the Mod- tain traces of burnt-out grass, although esty charcoal’s calibrated date is due to the lithic inclusions are comparable with the shape of the calibration curve at this those in the vessels. point, although the fact that at the 68.2% probability level of the calibrated The layer of ‘charred wood’ mentioned date range lies within 3370–3330 cal BC by Kinghorn is represented in the Muse- suggests that we are perhaps more like- um collection by five fragments of round- ly to be dealing with activity within the wood (Fig. 5.2), of which one (of Maloi- third, rather than the last, quarter of the deae species) provided the radiocarbon millennium. This date helps to dispel the date. That fragment, plus two of the oth- uncertainty about the chronology of fel- ers, was identified by Dr Susan Ramsey, site exploitation, which had hitherto re- who found that the others were of birch lied on observations about the discovery of (Betula). fragments of felsite knives or axeheads in later sites, e.g. at Stanydale (Calder Overall, the find from Modesty is intrigu- 1950), Tougs (Hedges 1986: 19 and see ing, and it raises the question of the na- p.30 for further examples) and Scord of ture of the deposit. The range of arte- Brouster (Ballin 2005: 15 and illus. 24. factual material present, and the Unfortunately, however, the provenance of condition of the axeheads and knives, this object within the Scord of Brouster 17

settlement was not recorded in detail: sels that can clearly be identified as Ballin and Whittle pers. comm.). How- Beaker domestic pottery (e.g. HD 930, ever, many more dates are required in HD 938 and HD 951a: Calder 1956: fig. order to establish the duration of felsite 17). Furthermore, a recent find of pottery exploitation, and these should emerge with herringbone-incised decoration from the planned fieldwork by Cooney et similar to that seen at Ness of Gruting, al. (as detailed in this volume). As had found at Ha’Breck on Wyre in Orkney, previously been noted (e.g. P. R. Ritchie has been dated to the end of the third/ 1968; 1992), virtually all of the products beginning of the second millennium BC of this exploitation remained in Shet- (Antonia Thomas pers. comm.). Further land, with only a very few reaching Ork- evidence indicating that the Ness of ney or mainland Scotland (P. R. Ritchie Gruting structure post-dates the Neo- 1992: 216; note that the Shetland knife lithic period is the set of four radiocar- purported to be from ‘Lanark’ is actually bon dates from the cache of carbonised from Lerwick: Fojut 2006). barley (of naked and hulled varieties), found inside the wall of the structure One misconception about activity at this (Table 1). Two of these dates have been time must be dismissed. It had previous- obtained by the Farming on the Edge ly been claimed that House 1 at Ness of Project; one was undertaken by the Gro- Gruting was of Neolithic date, on ac- ningen University laboratory in 1970, count of formal and decorative similari- having been submitted by Stuart Pig- ties between some of its pottery and gott, and the fourth was obtained from Hebridean Neolithic pottery (Calder the British Museum laboratory, having 1956: 356). Had these similarities been been submitted c. 1969 by R. B. K. Ste- borne out by the other evidence from venson of the (then-named) National that site, then by analogy with sites con- Museum of Antiquities of Scotland (Bar- taining dated Hebridean Neolithic pot- cham 1980). It is clear from these that tery (such as Eilean Domhnuill, Loch the Ness of Gruting structure was con- Olabhat, N. Uist: Armit 2003; cf. Sheridan structed c. 2200–2000 BC, and therefore 2008), it should date to the time-frame dates to what is, elsewhere in Britain, under investigation here. However, as termed the Early Bronze Age. As will be Audrey Henshall pointed out (Henshall clear from the discussion below, it may 1956, 383), the ceramic assemblage – be that many structures in Shetland that which appeared to form a unified whole, had initially been assumed to be of Neo- rather than the result of multiple phases lithic date are actually of Chalcolithic or of activity – also includes several ves- Bronze Age date.

Nordlige Verdener Shetlandsprojekt 18

Table 1: Radiocarbon dates for carbonised barley grains, Ness of Gruting, calibrated using OxCal 4.1.7 (using IntCal09, with atmospheric correction) and rounded to the nearest decade. Note: 1. the very large standard deviation in the BM-441 date reduces its usefulness. 2. GrN-6168 and BM-441 were determined before 13C values were routinely measured. 3. The dates can be combined, using OxCal 4.1.7, to produce a date of 3697±23 BP, 2140–2030 cal BC at 68.2% probability, 2200–1980 cal BC at 95.4% probability (see diagram).

Lab No. Barley type Date BP 13C Date cal BC, Date cal BC, When 68.2% 95.4% determined probability probability

AAR-15646 Hulled 3736±35 -20.8±0.6 2200–2050 2280–2030 2012

AAR-15647 Naked 3668±35 -21.6±0.6 2130–1980 2190–1950 2012

GrN-6168 Hulled & naked 3710±55 Not measured 2200–2030 2290–1950 1970

BM-441 [Presumably] 3514±120 Not measured 2020–1690 2200–1530 c1969 hulled & naked

Ness of Gruting R_Combine (3697,23)

3900 68.2% probability (17.8%) 2134-2113BC (50.4%) 2102-2037BC 95.4% probability 3800 (3.4%) 2194-2177BC (91.3%) 2145-2023BC (0.7%) 1991-1985BC

3700 Agreement 99.7%

3600

3500 Radiocarbon determination (BP)

3400

3300

2300 2200 2100 2000 1900 Modelled date (BC) 19

Among the many outstanding questions concerning this period of Shetland’s prehistory is that of what kind/s of funerary monument was/were being constructed (see below). In particular, the date at which the trefoil-chambered megalithic monuments (such as the Phase 1 structure at Vementry: Fig. 6) were constructed remains unknown, although the struc- tural history of that particular monument suggests that it pre-dated the practice of constructing heel-shaped cairns.

Developments in Shetland c. 3300- 3000 BC: Middle Neolithic Fig. 6: The passage tomb at Vementry, showing its original circular The picture becomes a little clearer at cairn and its (presumably) subsequent transformation into a heel- shaped cairn. After Henshall 1963, reproduced with permission of this point, since we know that the com- Edinburgh University Press. munal cist at Sumburgh Airport dates to this period (Hedges & Parry 1980; Melton 2008; Melton & Montgomery in press), shape, despite having flatter rims and as does the settlement activity pre-dat- being larger and coarser, and they also ing stone Houses 1 and 2 at Scord of share the use of grass as a filler with Brouster (Whittle et al. 1986, table 1) pottery from the Sumburgh cist (pot 2) and, perhaps, the earliest settlement and from Jarlshof (Hamilton 1956, 12-13 evidence at Jarlshof (Hamilton 1956, and fig. 6). As pointed out above, the 12-13; Dockrill et al. 2004) – but see be- calibrated date range for the Modesty low regarding its dating. Similar pottery, hoard extends into the period under con- featuring undecorated, thin-walled un- sideration here. As with the style of this carinated bowls, was found at all three pottery, the use of grass appears to be a sites (Fig. 7) and appears to represent a technical development peculiar to Shet- stylistic development local to Shetland. land, not being characteristic of Neolith- Pottery of similar shape occurs among ic pottery elsewhere in Scotland. the artefactual finds recovered on the Hill of Crooksetter. Furthermore the Melton and Montgomery’s work on the aforementioned vessels found at Mod- human remains from the Sumburgh cist esty are not dissimilar to these in overall has produced seven new radiocarbon

Nordlige Verdener Shetlandsprojekt 20

dates, in addition to the one published 2880–2490 cal BC (Whittle et al. 1986, by Hedges and Parry in 1980 (i.e. table 1; once again, no adjustment has 4395±55 BP from unburnt bone, 3330– been made to the standard deviation 2900 cal BC, GU-1075 – not applying the values). It would appear that CAR-243 is increase in standard deviation as rec- an outlier. The evidence pre-dating ommended by Ashmore et al. 2000 when House 2 (ibid., 5–8) consists of a dark dealing with old determinations). The soil layer with local concentrations of new dates appear to extend the overall ash and charcoal, together with possi- date span to c. 3500–3000 BC, although ble stakeholes and postholes suggest- this is no doubt due to the plateau in the ing an oval structure. Three radiocarbon calibration curve. Whether Bayesian dates (all from ‘birch etc.’ charcoal, one modelling could help to constrain the including charred barley) produced re- scatter here is unclear. Janet Mont- sults broadly comparable to those from gomery’s isotopic analysis of the human CAR-244–5 (namely 3370–2930, 3350– bones from the cist has, interestingly, 2930 and 3500–3020 cal BC, CAR-249– identified evidence for the consumption 251 respectively, ibid.: table 1). A date of marine resources (Melton & Mont- ostensibly relating to the construction of gomery 2009) and this may well consti- House 2 (CAR-252) is virtually identical tute the earliest dated example of this to these and some doubt must be cast dietary characteristic in the Scottish Ne- on whether it does actually date its con- olithic, contrasting with the picture ob- struction, since the House 2 ceramic as- tained for Neolithic human remains semblage looks to be Beaker domestic elsewhere in Scotland. (See Schulting pottery (ibid.: figs. 54, 55). et al. 2010 for the most recent review of this evidence.) The evidence from Jarlshof comes from the excavations undertaken by Gordon At Scord of Brouster, the evidence pre- Childe in 1937, where the pottery from dating House 1 (labelled as ‘phase 1’: Midden III and occupation layer III in- Whittle et al. 1986: 15–19) consists of cluded grass-tempered pottery which an anthropogenic layer including three was ‘quite distinct from the wares found hearths and a drainage gully. Three ra- in the later levels and houses and seems diocarbon dates – two from charcoal of to be a new variety with vaguely neo- ‘birch, etc.’, the third of birch – produced lithic affinities’ (Hamilton 1956, 13). two (CAR-244–5) that calibrate to 3350– Childe had remarked that this pottery 2930 and 3340–2710 cal BC respective- ‘has a soapy feel’ (Childe 1938: 360) – a ly, and one (CAR-243) that calibrates to characteristic shared with the Modesty 21

Fig. 7: Undecorated, uncarinated bowls of defi- nite and sus- pected Middle Neolithic date from the Sum- burgh cist (top), from Scord of Brouster (middle) and from Jarlshof (bottom).

From Hedges & Parry 1980; Whittle et al. 1986 and Hamilton 1956; various scales.

Nordlige Verdener Shetlandsprojekt 22

vessels. He also noted that some sherds Developments in Shetland c. 3000- appeared to have jabbed decoration. 2500 BC: Late Neolithic Further excavations undertaken by Ste- The presence of seven cushion mace- ve Dockrill and Julie Bond in 2004, de- heads in Shetland (Fig. 8.1) – a distinctive signed to elucidate the nature of agri- type of object, with a wide distribution cultural activities among these early in Britain but with concentrations in the inhabitants of Jarlshof, located what Northern Isles and the Thames valley was believed to be Childe’s Middens II (Gibson 1944; Roe 1968; 1979; Ritchie and III, and produced radiocarbon dates, 1992; Simpson & Ransom 1992: no. 58) – from charred barley grains, of 3455±35 indicates that Shetland was in contact BP (1880–1690 cal BC, GU-12916) and with areas to the south during this period. 3370±35 BP (1750–1540 cal BC, GU- That cushion maceheads date to the first 12915) for the lower midden (MIII) and half of the third millennium is indicated of 3260±35 BP (1610–1500 cal BC, GU- both by radiocarbon and by contextual 12914) for the upper midden (MII) (Dock- evidence. At Dorchester-on-Thames (site rill & Bond 2009: 50). Prima facie this II), Oxfordshire, an antler pick from the would appear to place the MIII activity primary fill of the ditch surrounding a within the Early Bronze Age; further dat- grave containing a short cushion mace- ing, of Childe’s bone finds from MIII, head produced a date of 4230±50 BP would be required in order to check this (2920-2630 cal BC, BM-4225N; Hey et al. since the pottery does not resemble Ear- 2011: 309 and fig. 12.42) while at Stone- ly Bronze Age pottery from Shetland. henge, another example that had been (This dating can be undertaken since the used as a grave good for a deposit of cre finds are in the National Museums Scot- mated bone is likely to date to between land collections.) 3030-2880 cal BC and 2570–2400 cal BC, to judge from Parker Pearson et al.’s Whether the felsite sources continued Bayesian-modelled dates for cremated to be exploited at this time remains to bone retrieved from Hawley’s excava- be demonstrated; the date from Modes- tions (Parker Pearson et al. 2009. The bone ty has also been discussed above. Fur- with which the macehead had been bur- thermore, the knowledge that a commu- ied is among an undifferentiated mass of nal cist was in use at this time raises cremated bone that had been excavated the question of whether chamber tombs from the ditch and the Aubrey Holes and were also being constructed or used; subsequently redeposited by Hawley in this must remain a research question for Aubrey Hole 7). Contextually, the discov- future excavation. ery of this artefact type at three Grooved 23

Fig. 8: Top: cushion macehead, Shetland (Shetland Museum, Reg. No. ARC 65223; 130 x 40 mm max dimensions); bottom: pestle macehead from Fleming- ton, Weisdale (Shetland Museum, Reg. No. ARC 74123, 107 x 67 mm). Pho- tos: Shetland Museum.

Ware-associated sites in Orkney – Ness with the fragment found at Barnhouse of Brodgar, Skara Brae and Barnhouse – being likely to belong towards the be- is compatible with a date within the over- ginning of this date bracket, and that all bracket c. 3100–2600 BC (extending found in the uppermost midden at Skara as late as c. 2300 BC at Ness of Brodgar), Brae likely to belong towards its end.

Nordlige Verdener Shetlandsprojekt 24

Incidentally, it should be noted that a While the maceheads indicate contact claim that a cushion macehead found at with the wider world, there is no evi- Knock, in Lewis may have been made dence that Shetland participated in the from riebeckite felsite, and thus had complex of activities associated with been an export from Shetland – a claim the use of Grooved Ware: there is noth- treated with caution by Roy Ritchie (P. R. ing in Shetland that remotely resembles Ritchie 1968: 132) – can now be dis- the henge at the Stones of Stenness, or counted, since thin-sectioning has the remarkable ‘temple complex’ at the shown it to be of calc-silicate hornfels Ness of Brodgar, for example. Indeed, from Creag na Caillich near Killin there is only one claimed example of (Sheridan 1992: 197). A second exam- Grooved Ware pottery in Shetland – a ple, found in Fife (NNMS X.AH 112), had large vessel with applied horizontal cor- also been claimed as a potential Shet- dons, represented by sherd 404 (and land export, and once again, Roy Ritchie perhaps also 147.1 and 535), from Sum- had expressed caution. This rock is burgh (Cleal & MacSween 1999: 203; clearly not of Creag na Caillich rock, nor Downes & Lamb 2000: fig. 27). Sherd does it macroscopically offer a close re- 404 had been found in a Late Bronze semblance to Shetland felsite; petrolog- Age/Early Iron Age house and was as- ical analysis would be required to deter- sumed to be redeposited. Whether this mine its identification and provenance. really is Grooved Ware, rather than a hitherto unrecognised form of first mil- The two pestle-shaped maceheads from lennium BC pottery, or redeposited pot- Shetland (Fig. 8.2) could date to this pe- tery from another period, is a moot point; riod as well – perhaps nearer to 3000 as the excavators pointed out, Shetland BC than to 2500 BC – since other exam- suffers from the lack of a clear ceramic ples from Late Neolithic contexts else- sequence (Downes & Lamb 2000: 60). It where are known, e.g. in the large pas- could also be noted that while the Ness sage tomb at Knowth, Ireland. However, an of Gruting assemblage includes some additional, much later use of this form sherds with a passing resemblance to of macehead is attested by the minia- Grooved Ware (sherd NMS X.HD 945), ture example found in a child’s grave, the Early Bronze Age date of that as- associated with a Food Vessel, at Doune, semblage now seems unimpeachable. Perthshire; by analogy with a miniature battle axehead found in another child’s Other evidence for activities during the grave nearby, this may well date to c. first half of the third millennium comes 1800-1600 BC (McLaren 2004). from Scord of Brouster, where the radio- 25

carbon dates suggest that House 1 could thy of mention here, since it is likely that have been built at this time (Whittle et many of the houses, enclosures and field al. 1986: table 1). If that is the case, then systems formerly assumed to be Neo- the associated assemblage of large, un- lithic date are actually of this date, or decorated, globular and carinated jars later. That Shetland was exposed to at and flat-based pots (ibid.: figs. 56 and least some of the Continental novelties 57) constitutes a distinctively Shetland that characterise the Chalcolithic period style of pottery, with no obvious exter- – even if no copper or early gold arte- nal comparanda. facts have been found – is indicated by a weathered sherd of international style, As regards funerary monuments, the ab- All-Over-Cord decorated Beaker from sence of dating evidence for Shetland’s Stanydale. Here, the novel pottery style chamber tombs makes it impossible to seems to have been used in a structure say whether any were constructed or wholly of local Shetland design. A new used at this time. At Scord of Brouster, a style of funerary monument is also indi- funerary monument which was de- cated by the cist with the inhumed skele scribed by the excavators as a kerb- ton of an adult male and Beaker at Fraga, cairn, but which actually looks to be a Scatness (Bryce 1932), which echoes ring-cairn with an internal cairn (Whittle Beaker funerary practice on ‘mainland’ et al. 1986: 36–43), produced a radiocar- Scotland. And the small but growing bon date of 4220±75 BP (3010–2580 cal number of assemblages of domestic BC, CAR-242) from ?birch charcoal from Beaker, from Stanydale, Ness of Gruting, a basal deposit. However, whether this Scord of Brouster House 2, the timber charcoal actually related to the con- structure at Sumburgh (Downes & Lamb struction or use of this monument is un- 2000) and Tougs (Hedges 1986), indi- clear; unfortunately, the scraps of cre- cates that the ceramic tradition took root, mated bone found with a pot rim in the and developed along its own trajectory. inner cairn are probably too small to be The two radiocarbon dates relating to radiocarbon-dated. the use of the oval house and associated burnt mound at Tougs (Hedges 1986:12) Subsequent developments in are comparable with those obtained for Shetland: Chalcolithic and Early the Ness of Gruting house. Bronze Age, c. 2500–1800 BC What happened after c. 2500 BC – when, Furthermore, it is evident that steatite on ‘mainland’ Britain, metal and other vessels were in use during this period, Continental novelties appeared – is wor- since a steatite cinerary urn found at

Nordlige Verdener Shetlandsprojekt 26

Fig. 9: Bone object from Jarlshof with Beaker- like decoration. Length: 127 mm; maximum width: 38 mm. Photo: National Museums Scotland.

Quandale in Orkney – which must have Secondly, the question of whether cham- been exported from Shetland – has been ber tombs continued to be built within dated (through its associated cremated the 2500–1800 BC period – and whether bone) to 3660±50 BP (2200–1900 cal BC, it is to this period that the heel-shaped GrA-19989). (and square) cairns were constructed – needs to be investigated through exca- Several issues remain to be resolved, vation. (Heel-shaped cairns are not however. Firstly, whether the enigmatic unique to Shetland, with examples bone object from Jarlshof with its Beak- known from Caithness, Davidson & Hen- er-like geometric design (Fig. 9) formed shall 1991: 41-2.) The parallelism be- part of the Beaker phenomenon in Shet- tween the concave façade of heel- land – and what it actually was – re- shaped cairns and the concave façade main a moot point. It had been found by of the large structure at Stanydale could Childe in Midden II; as noted above, be taken to imply possible contempora- charred barley grain/s from material be- neity, and the presence of a possible lieved to correspond to MII was/were Beaker sherd at Giant’s Grave may indi- radiocarbon-dated to around the middle cate that that monument had already of the second millennium: see GU-12914 been built, or was constructed around above, and Dockrill & Bond 2009: 50. the same time as the pot’s use. The evi- However, the only way to resolve the dence from Vementry suggests that the question of the artefact’s date is through heel-shaped cairn represents a re-shap- direct dating, and given its thinness, it ing of a pre-existing round cairn, thereby would be difficult to sample without indicating a ‘round-to-heel-shaped’ se- risking damage to the object. quence. Furthermore, the Muckle Heog 27

monument (Henshall 1963) – which fea- Thirdly, as before, the question of wheth- tures the interment of unburnt bodies er felsite artefacts were still being man- accompanied by steatite vessels within ufactured at this time needs to be re- cists under a heel-shaped cairn – may solved; was it that people were simply well date to the period 2200–1800 BC, re-using ancient objects? since we know from the Quandale evidence that steatite vessels were being Subsequent developments, made at this time. (As will be seen be- c 1800-1500 BC low, however, the use of steatite ves- Although, strictly speaking, these fall sels in the Northern Isles continued af- outside the scope of this paper, certain ter 1800 BC; see also Foster & Turner observations can usefully be offered here. 2009 on the use of steatite.) The Muckle Heog monument is particularly interest- Firstly, regarding settlements and ce- ing as it appears to represent a fusion of ramic typochronology, continuity in the traditional Shetland practice (i.e. of con- general style of house construction (i.e. structing cairns for funerary monuments) thick-walled and cellular) is suggested with novel practices (i.e. the use of cist by the evidence from Scord of Brouster, graves and of steatite vessels), with the where House 3 was constructed at this practice of using cists probably being time. The radiocarbon dating of organic adopted from outside Shetland. Wheth- residue adhering to a large undecorated er this site also marks the end of a short- jar of steatitic clay from the Bunyie (Be- lived period of heel-shaped cairn con- nie) Hoose, Whalsay (Calder 1961) – un- struction remains to be demonstrated dertaken as part of National Museums’ through excavation of other sites with Scotland radiocarbon dating programme heel-shaped cairns. In any case, Muckle (Sheridan 2005) – has revealed that this Heog – like the (probably slightly earlier) house was in use during the second Beaker cist at Fraga – reminds us that, quarter of the second millennium, al- during the period under consideration though whether it was built then, or be- here, the inhabitants of Shetland prob- forehand, is uncertain. Like Stanydale, it ably had a choice of funerary practice has a concave façade. The organic resi- (i.e. interment in a passage tomb, vs. a due produced a date of 3360±40 BP cist). This may relate to the exercise of (1740–1530 cal BC, GrA-29373: Sheridan choice as to whether to maintain a long- 2005: 183) and is useful for indicating standing local tradition, or to adopt new the post-Beaker trajectory of ceramic fashions from elsewhere in Scotland. development in Shetland.

Nordlige Verdener Shetlandsprojekt 28

Secondly, as regards funerary practices, unfinished miniatures in the Ness of the same NMS radiocarbon dating pro- Gruting house therefore suggests some gramme has shown that the tradition of time-depth to the use of this structure. cremation and deposition of the remains in a cinerary urn – a practice in wide- Conclusions spread use elsewhere in Scotland – was From the foregoing it is clear that our definitely in use in Shetland by the first knowledge and understanding of the quarter of the second millennium, with Shetland Neolithic leaves a great deal cremated bone from an urn of steatitic to be desired. While we can begin to clay from Culla Voe producing a date of sketch an overall narrative, as attempt- 3475±40 BP (1890–1690 cal BC, GrA- ed here, the gaps in our knowledge are 24056: Sheridan 2007b: 184). The possi- substantial and much targeted fieldwork bility that the Muckle Heog monument, will be necessary to address the many with its steatite vessels, might also date outstanding questions. to this period has also been mentioned above; we know that steatite urns con- However, it is clear that a particularly in- tinued to be exported to Orkney during teresting story is waiting to be told – of this period (ibid.). episodes of contact with the outside world (but not with Scandinavia!), inter- Finally, the unfinished miniature battle- spersed with periods when a strong axeheads found at the Ness of Gruting Shetlandic insular identity (as expressed, (together with a finished example found for example, in architecture and materi- in the Sumburgh settlement: Downes & al culture) developed. It may be that the Lamb 2000: 67 and fig. 29) indicate practice of farming, and the beliefs, awareness of, and participation in, the practices and traditions that accompa- ‘vocabulary of esteem’ that was current nied this, arrived in Shetland from west- during the early second millennium BC ern Scotland during the late 38th or 37th elsewhere in Britain. A similar miniature century BC; further excavation at West battle-axehead, found in a child’s grave Voe, and investigation of the simplest along with a Food Vessel, in Doune, passage tombs (including Ronas Hill) Perth & Kinross, has been radiocarbon might help to clarify this. How these pu- dated to 3400±35 BP (1870–1610 cal tative incomers interacted with the in- BC, SUERC-2869: Sheridan 2007b: 185) digenous fisher-gatherer-hunter groups and this is consistent with the dating of (as attested at the lowest level of West full-sized versions of similar shape (ibid.: Voe) is unknown, although the 37th cen- 175 and fig. 14.10). The presence of two tury activity at West Voe could conceiv- 29

Fig. 10: Notional sequence of funerary monument types in Shetland. Sites featured, from left to right and top to bottom: Ronas Hill; Vord Hill, South; Vementry (phase 1); the Sumburgh cist; Seli Voe; Vementry (phase 2); Punds Water; March Cairn; Pettigarths Field (chamber tomb); Pettigarths Field (cist); Muckle Heog, West.Excepting the Sumburgh cist, the dating of these monuments is speculative. Site plans after Henshall 1963, Calder 1961 and Hedges & Parry 1980.

Nordlige Verdener Shetlandsprojekt 30

ably relate to a process of acculturation, tion of the practice of individual cist in- whereby elements of the new lifestyle terment represents the emulation of a (principally pottery use) had been adopt- Beaker-associated practice elsewhere ed by one such group. Thereafter, for in Scotland. However, the development over half a millennium, a process of in- of Shetland Beaker pottery is unmistak- sular development took place, as attest- ably localised. ed most clearly in the emergence of a Shetland style of Neolithic pottery and The review of the domestic evidence as in the use of felsite to create distinctive presented in this contribution indicates knives and axeheads – the largest of that there are very few settlements that which can only have been non-utilitari- pre-date 2500 BC, in contrast to the pean, given the paucity of trees on the ar- riod 2500–1500 BC; if this reflects a chipelago. The next evidence suggest- genuine pattern, then we may be wit- ing contact with the outside world nessing a process of population growth – assuming that the north-east Irish por- from a small Neolithic base. Under- cellanite axeheads mentioned above standing the subsistence basis and so- had arrived as part of the initial Neoli- cial dynamics that could have under- thisation process, and in the absence of pinned such a development will require dating evidence for the few felsite ex- more excavation and palaeoenviron- ports from Shetland – is not until the mental work of the type undertaken by first half of the third millennium BC, be- Dockrill and Bond, Edwards et al. and ing reflected in the use of symbols of Melton et al.; pointers as to areas of power that have a wide distribution in high potential already exist in the re- Britain. Their adoption may well reflect sults of field and aerial survey, as men- contact with Orkney, a link that recurs at tioned above. various times thereafter; but clearly Shetland did not partake in the competi- In terms of funerary practices, although tive conspicuous consumption that a sequence as sketched in Fig. 10 can be marks out Late Neolithic society in Ork- proposed, it remains to be tested through ney (as seen, for example, in the com- excavation and dating. Once more, a dy- plex of sites and monuments around the namic between a process of localisation Loch of Stenness). Other contact with and the adoption of exogenous practic- the external world, but probably not via es can be traced. Orkney, is attested by the presence of All-Over-Cord Beaker, albeit occurring in As for the question of felsite exploita- a Shetland-style context; and the adoption (and indeed the use of other lithics), 31

it is anticipated that the research Bibliography planned by Cooney et al. should clarify its chronology and nature. Armit, I. 2003: The Drowners: perma- nence and transience in the Hebridean There is therefore much still to do, but Neolithic. Neolithic Settlement in Ire- Shetland constitutes one of the most land and Western Britain. Armit, I. Mur- exciting areas in which to undertake Ne- phy, E. Nelis, E. & Simpson, D. D. A. olithic research, given the excellent (eds.), Oxbow, Oxford: 93-100. state of preservation of buildings and land divisions. It is hoped that the Farm- Ashmore, P. J., Cook, G. T. & Harkness, ing on the Edge: Cultural Landscapes of D. D. 2000: A radiocarbon database for the North Project will prove to be a cata- Scottish archaeological samples. Radio- lyst for the fieldwork and analysis that carbon. Vol. 42: 41-48. needs to be done. Ballin, T. B. 2011a: The Post-Glacial col- Acknowledgements onization of Shetland – integration or Ditlev Mahler is warmly thanked for in- isolation? Evidence from lithic and stone viting me to participate in his exciting assemblages. Farming on the Edge: Cul- project, and for his patience as an edi- tural Landscapes of the North. Mahler, tor. Ian Tait is thanked for his help with D. L. & Andersen, C. (eds.), National Mu- macehead images; Rick Schulting is seum of Denmark, Copenhagen: 32-43. thanked for his help in processing the Ness of Gruting dates; and copyright Ballin, T. B. 2011b: The felsite quarry holders are thanked for allowing me to complex of Northmaven: observations reproduce illustrations. Craig Angus pro- from a fact-finding mission to Shetland. vided invaluable assistance with pre- Flint and Stone in the Neolithic Period. paring the illustrations. Saville, A. (ed.), Oxbow, Oxford: 62-81.

Nordlige Verdener Shetlandsprojekt 32

Ballin, T. B. 2011c: Hill of Crooksetter, Calder, C. S. T. 1961: Excavations in Delting, Shetland. The lithic assem- Whalsay, Shetland, 1954–5. Proceed- blage. Unpublished specialist report for ings of the Society of Antiquaries of ORCA. Scotland. Vol. 94 (1960-61): 28-45.

Barcham, R. C. 1980: A lost radiocarbon Calder, C. S. T. 1963: Cairns, Neolithic date for Shetland. Proceedings of the houses and burnt mounds in Shetland. Society of Antiquaries of Scotland. Vol. Proceedings of the Society of Antiquaries 110 (1978-80): 502-506. of Scotland. Vol. 96 (1962-63): 37-86.

Brend, A. 2010: Laggan-Tormore-Crook- Cleal, R. & MacSween, A. 1999: Grooved setter Hill. Discovery and Excavation in Ware in Britain and Ireland. Oxbow, Oxford. Scotland. Vol. 11: 157-8. Cumming, G. A. 1946: Flint core axe Brend, A. & Barton, R. 2011: Hill of Crook found on Fair Isle, Shetland. Proceed- setter, Delting, Shetland. Excavation at ings of the Society of Antiquaries of Site 002. Unpublished Data Structure Scotland. Vol. 80 (1945-46): 146-148. Report, Project 246, ORCA, Kirkwall. Davidson, J. L. & Henshall, A. S. 1989: Bryce, T. H. 1932: Notice of a short cist The Chambered Cairns of Orkney. Edin- at Fraga, Scatness, Shetland. Proceed- burgh University Press, Edinburgh. ings of the Society of Antiquaries of Scotland. Vol. 67 (1932–3): 34-36. Davidson, J. L. & Henshall, A. S. 1991: The Chambered Cairns of Caithness. Ed- Calder, C. S. T. 1950: Report on the exca- inburgh University Press, Edinburgh. vation of a Neolithic temple at Stany- dale in the parish of Sandsting, Shet- Dockrill, S.J., Bond, J. & Batey, C.E. land. Proceedings of the Society of 2004: Jarlshof (Dunrossness parish), Antiquaries of Scotland. Vol. 84 (1949- multi-period settlement. Discovery and 50): 185-205. Excavation in Scotland. Vol. 5: 116-7.

Calder, C.S.T. 1956: Report on the dis- Edwards, K. J., Schofield, J. E., Whitting- covery of numerous Stone Age house- ton, G. & Melton, N. D. 2009: Palynology sites in Shetland. Proceedings of the So- ‘on the edge’ and the archaeological ciety of Antiquaries of Scotland. Vol. 89 vindication of a Mesolithic presence? (1955-56): 340-397. The case of Shetland. Mesolithic Hori- 33

zons. Papers Presented at the Seventh Hedges, J. W. 1986: Bronze Age struc- International Conference on the Meso- tures at Tougs, Burra Isle, Shetland. lithic in Europe, Belfast 2005. Volume 1. Glasgow Archaeological Journal. Vol. McCartan, S. Schulting, R. J., Warren 13: 1-43. G.& Woodman P. (eds.), Oxford and Oakville, Oxbow: 113-123. Hedges, J. W. & Parry, G. A. 1980: A Ne- olithic multiple burial from Sumburgh, Fojut, N. 2006: Prehistoric and Viking Shetland. Glasgow Archaeological Jour- Shetland. Shetland Times, Lerwick. nal. Vol. 7: 15-26.

Foster, A. K. & Turner, V. E. 2009: Kleber: Henshall, A.S. 1956: Appendix II. Pottery Shetland’s Oldest Industry. Soapstone and stone implements from the Ness of since Prehistory. Shetland Heritage Pub- Gruting. In C. S. T.Calder, Report on the lications, Lerwick. discovery of numerous Stone Age house- sites in Shetland, 381-391. Proceedings Gibson, W. J. 1944: Maceheads of of the Society of Antiquaries of Scot- ‘cushion’ type in Britain. Proceedings of land. Vol. 89 (1955-56): 340-397. the Society of Antiquaries of Scotland. Vol. 78 (1943-44): 16-25. Henshall, A. S. 1963: The Chambered Tombs of Scotland. Volume 1. Edinburgh Gilmore, G. K. & Melton, N. J. 2011: Early University Press, Edinburgh. Neolithic sands at West Voe, Shetland Islands: implications for human settlement. Henshall, A. S. 1972: The Chambered Human Interaction with the Geosphere: Tombs of Scotland. Volume 2. Edinburgh The Geoarchaeological Perspective. Wil- University Press, Edinburgh. son, L. (ed.) Geological Society of London (Special Publications, 352), London: 69-83. Hey, G., Garwood, P., Robinson, M., Bar- clay, A. & Bradley, P. 2011: Thames Hamilton, J. R. C. 1956: Excavations at Through Time. The Archaeology of the Jarlshof, Shetland. Her Majesty’s Sta- Gravel Terraces of the Upper and Middle tionery Office (Ministry of Works- Ar Thames. Part 2 – Mesolithic to Early chaeological Reports No. 1), Edinburgh. Bronze Age. Oxford Archaeology, Thames Valley Landscapes Monograph No. 32, Hedges, J. W. 1984: Gordon Parry’s Oxford. West Burra survey. Glasgow Archaeo- logical Journal. Vol. 11: 41-59.

Nordlige Verdener Shetlandsprojekt 34

Jóhansen, J. 1976: The age of the intro- land’s earliest known archaeological site. duction of Plantago lanceolata to the The New Shetlander. Vol. 233: 21-22. Shetland Islands. Danmarks Geologiscke Undersogelse Arbog 1976: 45-48. Melton, N. D. 2008: West Voe: A Meso- lithic-Neolithic transition site in Shet- Jóhansen, J. 1985: Studies in the Vege- land. Scottish Odysseys: the Archaeology tational History of the Faroe and Shetland of Islands. Noble, G., Poller, T. Raven J. & Islands. Tórshaven, Annales Societatis Verrill, L. (eds.), Tempus, Stroud: 23-36. Scientiarum, Faeroensis Supplementum 11. Melton, N. D. 2009: Shells, seals and Kinghorn, G. 1895: Notes on the discovery ceramics: an evaluation of a midden at of a deposit of polished stone axes and West Voe, Sumburgh, Shetland, 2004– oval knives of porphyry, &c., at Modesty, 5. Mesolithic Horizons. Papers Present- near Bridge of Walls, Shetland. Proceed- ed at the Seventh International Confer- ings of the Society of Antiquaries of ence on the Mesolithic in Europe, Belfast Scotland. Vol. 29 (1894-95): 48-54. 2005. Volume 1. McCartan, S. Schulting, R. J., Warren G.& Woodman P. (eds.), McLaren, D. 2004: An important child’s Oxford and Oakville, Oxbow: 184-189. burial from Doune, Perthshire, Scotland. From Sickles to Circles: Britain and Ire- Melton, N. D. 2010: Lerwick harbour. land at the Time of Stonehenge. Gibson, Discovery and Excavation in Scotland. A. M. & Sheridan, J. A. (eds.), Tempus, Vol. 11: 158. Stroud: 289-303. Melton, N.D. & Montgomery, J. 2009: Mahler, D. L. 2010: http://nordligever- Sumburgh Early Neolithic burial cist: com dener.natmus.dk/uploads/media/Shet- bined isotope analysis. Discovery and land_report_2010.pdf Excavation in Scotland. Vol. 10: 166.

Mahler, D. L. & Andersen, C. (eds.), 2011: Melton, N. D. & Nicholson, R. A. 2004: Farming on the Edge: Cultural Land- The Mesolithic in the Northern Isles: the scapes of the North, National Museum preliminary evaluation of an oyster mid- of Denmark, Copenhagen. den at West Voe, Sumburgh, Shetland, UK. Antiquity. Vol. 78, Project Gallery: Melton, N. D. 2005: Investigations at http://antiquity.ac.uk/ProjGall/nichol- West Voe, Sumburgh: new work at Shet- son/index.html 35

RCAHMS (Royal Commission on the An- Roe, F. E. S. 1979: Typology of stone im- cient and Historical Monuments of Scot- plements with shaftholes. Stone Axe land), 2010: Royal Commission on the Studies. Clough, T. H. McK. & Cummins, Ancient and Historical Monuments of W. A. (eds.), Council for British Archaeol- Scotland. Discovery and Excavation in ogy (Research Report 23), London: 23-48. Scotland. Vol. 11: 193-204. Saville, A. 1994: A possible Mesolithic Ritchie, J. N. G. 1970: Excavation of the stone axehead from Scotland. Lithics. chambered cairn at Achnacreebeag. Pro Vol. 15: 25-28. ceedings of the Society of Antiquaries of Scotland. Vol. 102 (1969-70): 31-55. Saville, A. 2000: Orkney and Scotland be fore the Neolithic period. Neolithic Orkney Ritchie, P. R. 1968: The stone implement in its European Context. Ritchie, A. (ed.), trade in third-millennium Scotland. Stud- Cambridge, McDonald Institute: 91-100. ies in Ancient Europe. Essays Presented to Stuart Piggott. Coles, J. M. & Simp- Schulting, R. J., Sheridan, J. A., Crozier, son, D. D. A. (eds.), Leicester University R. & Murphy, E. 2010: Revisiting Quanter- Press, Leicester: 117-136. ness: new AMS dates and stable isotope data from an Orcadian chamber Ritchie, P. R. 1992: Stone axeheads and tomb. Proceedings of the Society of An- cushion maceheads from Orkney and tiquaries of Scotland. Vol. 140: 1-50. Shetland: some similarities and con- trasts. Vessels for the Ancestors: Essays Sheridan, J.A. 1986: Porcellanite arte- on the Neolithic of Britain and Ireland in facts: a new survey. Ulster Journal of Honour of Audrey Henshall. Sharples, N. Archaeology. Vol. 49:19-32. M. & Sheridan, J. A. (eds.), Edinburgh University Press, Edinburgh: 213-220. Sheridan, J. A. 1992: Scottish stone axeheads: some new work and recent dis- Roe, F. E. S. 1968: Stone mace-heads coveries. Vessels for the Ancestors: Es- and the latest neolithic cultures of the says on the Neolithic of Britain and British Isles. Studies in Ancient Europe. Ireland in Honour of Audrey Henshall. Essays Presented to Stuart Piggott. Sharples, N. M. & Sheridan, J. A. (eds.), Coles, J. M. & Simpson, D. D. A. (eds.), Edinburgh University Press, Edinburgh: Leicester University Press, Leicester: 194-212. 145-172.

Nordlige Verdener Shetlandsprojekt 36

Sheridan, J.A. 2005: The National Mu- F. Healy & A. Bayliss, Gathering Time: seums’ of Scotland radiocarbon dating Dating the Early Neolithic Enclosures of programmes: results obtained during Southern Britain and Ireland. Antiquity. 2004/5. Discovery and Excavation in Vol. 86 (331): 262-264. Scotland 6, 182–3. Simpson, D. D. A. & Ransom, R. 1992: Sheridan, J. A. 2007a: From Picardie to Maceheads and the Orcadian Neolithic. Pickering and Pencraig Hill? New infor- Vessels for the Ancestors: Essays on the mation on the ‘Carinated Bowl Neolithic’ Neolithic of Britain and Ireland in Honour in northern Britain. Going Over: the Me- of Audrey Henshall. Sharples, N. M. & solithic–Neolithic Transition in North- Sheridan, J. A. (eds.), Edinburgh Uni- West Europe. Whittle, A.W.R. & Cum- versity Press, Edinburgh: 221-243. mings, V. (eds.), Oxford University Press (Proceedings of the British Academy. Turner, V. 2011: From homestead enclo- Vol. 144), Oxford: 441-492. sure to farm? Field development in Shet- land in the Neolithic period. Farming on Sheridan, J. A. 2007b: Dating the Scot- the Edge: Cultural Landscapes of the tish Bronze Age: “There is clearly much North. Mahler, D. L. & Andersen, C. (eds.), that the material can still tell us”. Be- National Museum of Denmark, Copen- yond Stonehenge: Essays on the Bronze hagen: 19-31. Age in Honour of Colin Burgess, Bur- gess, C. Topping, P. & Lynch, F. (eds.), Ox- Whittle, A. W. R., Keith-Lucas, M., bow, Oxford: 162–85. Milles, A., Noddle, B., Rees, S. & Ro- mans, J. C. C. 1986: Scord of Brouster. Sheridan, J. A. 2008: Pottery from Barpa An Early Agricultural Settlement on Langais. Unpublished specialist report Shetland. Oxford University Committee prepared for ARCUS. for Archaeology, Oxford.

Sheridan, J. A. 2010: The Neolithisation Whittington, G. 1978: A sub-peat dyke of Britain and Ireland: the Big Picture. on Shurton Hill, Mainland, Shetland. Landscapes in Transition. Finlayson, B. & Proceedings of the Society of Antiquaries Warren, G. (eds.), Oxbow, Oxford: 89-105. of Scotland. Vol. 109 (1977-78), 30-35. Sheridan, J. A. 2012: Review of A. Whittle, 37

Pinhoulland – a multi period site, West Mainland, Shetland

Ditlev L. Mahler Fig. 1: Pinhoulland seen from Browland East of the voe.

Photo Ditlev L. Mahler 2011.

Nordlige Verdener Shetlandsprojekt 38

General introduction vide us with a whole new perspective The object of the three years research on the Neolithic process in the area and project Shetland – the Border of Farming provide comparative material with re- 4000-3000 B. C. E. is to gather, analyse gards to Scandinavia. In the light of the and document the Shetland Island’s Ne- well documented, early Neolithic South- olithic material in order to deepen our ern Scandinavian material it should be understanding of the Neolithic process possible to create a chronological order and the social impact on then existing as a basis for an understanding of the societies within the period, both on the Neolithic process in general and on the Shetland Islands as well as in Scandina- Shetland Islands specifically. via. Newer research suggests that the Shetland Islands became Neolithic by With regards to the expansion of agrar- several steps over a period of time ian societies, the Shetland Islands were (Mahler 2011: 13). The research project the ultimate North Western boundary aims to survey the sequence and nature lying 60° northern latitude. Within the of the various elements in this process time span 4.000-3.000 B.C.E an agrarian as a basis for a comparative analysis of culture established itself on the Shet- the Scandinavian agrarian societies and land Islands with material and immate- their expansion. rial elements such as ritual practices characteristic of the Neolithic process in The project also applies landscape anal- all of Europe. Amongst these elements ysis in order to expose the ritual land- are monumental sites such as the cham- scape as well as cultural markers in the bered cairn at Caldback, Unst, which same landscape. The project could pro- probably is the Northernmost passage grave of Europe still in existence. And Photo Ditlev L. Mahler 2011. also polished tools of ritual use such as point butted felsit axes and geometri- cally ornamented ceramic ware. A relatively large population established, at an early date, stone boundaries around their fields and across the landscape in

Fig. 2: The chambered cairn at Caldback, Unst, seen from South East. The passage grave could be the Northern most passage grave in Europe still in existence. 39

general which could be interpreted as but it is still an expansive achievement. territorial boundaries, and there are ex- Shetland was not a virgin land when the amples that long stone dykes were build Neolithic people arrived (Edwards et al. for ritual causes at Vemmentry below 2009: 118 f.; Melton 2009: 188; Gilmore the cairn to the East and at Lardie Hill & Melton 2011: 80). We cannot be sure South of Stanydale. In both cases the that there were a stable Mesolithic pop- dykes separate the land of the living ulation, but it is highly likely that Shet- from the land of the dead. At Stanydale land had Mesolithic visits (Bennet 1992: a hall like construction has been uncov- 241). It would be strange if Shetland ered, a possible parallel to the Conti- was the only place in Northern Europe, nental gathering sites, and Surveys dur- which did not have a Mesolithic popula- ing 2010 and 2011 and mapping of the tion at least during periods. It is unlikely site suggest that the site probably was though that there was a Mesolithic pop- something special and contains a more ulation, who went Neolithic (compare complicated relative chronology (Calder Sheridan 2010: 93, 101). 1951: 185 ff.; 1956; 1964). This summer’s six week field work led to Another approach is the comparative the visit of 12 Neolithic sites mainly in study and models for island societies West Mainland and some sites on Unst. and for example their demographic and Of these 12 sites six were mapped using economic development compared with precision GPS among other sites we the human intervention in virgin nature. mapped Pinhoulland, which will be the Such a comparison can be based on pio- topic of this article. neer societies with or without written sources on their development. An exam- Pinhoulland – an introduction ple is the Lapita dispersal in Oceania or The site has been known for a long time, Sarqqaq Culture dispersal in Greenland so that is not the reason for mentioning both a. 2000 B. C. E. (Mahler IP.; Kirch Pinhoulland once again. One unroofed 1988:103 ff.; Grønnow 2004: 78; Grøn- structure is thus depicted on the first now IP.). The reason the Shetland Is- edition of the OS 6-inch map from 1882, lands were the northern boundary for an and the survey from 1968 mentions nine agrarian expansion could possibly be oval houses, enclosures and a large field due to a maritime technological devel- system. Already at this time it must have opment. Seen from a maritime aspect been clear that Pinhoulland represented the Orkney Islands and Fair Isle are step- one of the largest prehistoric settlements ping stones in the Neolithic expansion of Neolithic/Bronze Age nature on Shet-

Nordlige Verdener Shetlandsprojekt 40

land. Whittle (1979: 167, 1986: 53f, Fig. centrate on a minor part of the cultural 47 & 49) has an interesting map showing landscape and chose the Eastern slope seven to nine house sites or structures for mapping. The site consists of rather at Pinhoullend, a house site at South acidic grassland with areas of very wet Stany Fields further North besides Scord and peaty ground being the remains of of Brouster. Together with the house sites now fossile small water courses. In all at Loch of Grunnavoe, the map illustrates we mapped 360 x 215 m2 during the five very well the extensive prehistoric settle days spent there in 2011. At the time ment systems in this part of West Main- before the growth of the blanket peat, land, though it is obvious that the pic- Pinhoulland must have been a very suit- ture must be diachronic. It is also very able area for agriculture. Originally the understandable that many archaeologists area must thus have been well drained have been fascinated by the site (Turner representing a nice agricultural poten- 2011: 22, Fig. 2b), and the following de- tial. In this light the many structures and scription and interpretation of Pinhoul- house sites are not surprising (compare land must be seen as one in a long row Edwards & Whittinton 1998: 17). to be followed, and only massive excavations will reveal the actual nature and In the following I shall try laying down chronology of the site. This ought not to the questions and doubts of the inter- prevent us from having a look at our pretation of the structures. I willingly present knowledge of Pinhoulland. admit that many of the interpretations invite discussion and fruitful disagree- The first time I visited Pinhoulland was ments, and this will continue as long as during summer 2010 approaching the site no excavations have been carried out. from the coast line of Voe of Browland. We have very few excavations as a At first glance the many structures were whole of the Neolithic structures on quite overwhelming, and beside a krobb Shetland – C.S.T. Calder’s post-war in- and the 19th Century pen down at the vestigations, then a long pause before coast, the site seems rather undisturbed Alasdair Whittle’s excavation during which of course is deceiving. Most of 1979 and then recent investigations at the structures are placed on the gentle Sullom Voe – not yet published (Fojut Eastward slope down to Voe of Brow- 2006). This also means that we may ex- land, but dykes, structures and clear- pect many surprises about the nature of ance cairns continue on the other side Shetland’s Neolithic Period, and with of the hill as implied above. It was clear Orkney and Caithness in mind we must from the beginning that we had to con- keep an open mind. 41

Fig. 3: Map of the area around Pinhoulland and Scord of Brouster. After Whittle 1986, fig. 47.

Nordlige Verdener Shetlandsprojekt 42 ± ±

1 15 14 1 16 18 15 14 13 ? 16 18 17 13 ? 12 17 12 10

10 7 9

7 9 Fig. 4: Over view of Pinhoulland as 11 mapped during field work in 2011. 11 Brown: Structures /house sites; 2 6 red: Clearance cairns; grey: Dykes; 2 6 green: The two banks. Mapped by M. Hoydal/ processed by P. Jensen.

19 8 5 19 3 5 8 0 25 50 100 4 3 Meters 0 25 50 100 4 Meters The site Alasdair Whittle has of course also the large mound, which in the following mapped the site (1986, fig. 49), but will be one of my focus points. The many strangely enough he did not discuss the clearance cairns indicate that the area interpretation of the more dubious struc- must have been intensively farmed either tures or especially the large mound to for grazing or for producing barley. Many the North (1986: 54 f.). Fig. 4 shows the of the cairns measure several meters in area mapped this summer, especially diameter. 43

Most of the house structures are rather House 10 lies beneath a krobb and is clear, and in at least two cases it is pos- rather dilapidated because of being sible to distinguish several phases. robbed of stones; the same goes for House 1 has been rebuilt at least once structure 12, but both 10 and 12 are like- but has kept the same entrance towards ly to be house structures. In front of the South or South East. The small house house 12, to the North West, there is a 6 which measures about 6, 5 m internal- standing stone; one of the two ortholits ly has been built on an older and larger (standing stones) at Pinhoulland. One of structure measuring about 10, 5 m in the most central house structures is length. Structure 11 could very well be house 9 as the connected dykes bind the remains of an older house site. On many of the other structures together the other hand, the relationship be- with house 9, and the dykes could have tween structure 5 and 8 are unclear. We chronological implications. could be facing two house structures as is the case with house 6 and 11. On the The questionmark South of house 1 other hand both structures consist of marks an oval structure measuring rather large stones tumbled together, around 8 m and with protruding stones which could be the remains of one or indicating that the small mound must be two cairns. manmade and may be covering a house structure. East of the mound and some- In connection with his map (1986: fig. what lower down the slope there is a 49) Whittle does not interpret what he cluster of houses surrounded by dykes calls turf banks, and I call structure 3, 4 and two banks. Five of the houses are and 19. Structure 4 could be a pen built built together, and at least four have en- on to the dyke running towards the Loch trances pointing to the South, towards of Grunnavoe, but the two other struc- house structure 17, where protruding tures are more likely to be house sites. stones indicate an entrance pointing to Structure 4 could even be of much the North. This gives an impression of younger date. The structures 3, 4, 5, 8 the existence of a kind of passage be- and 19 are all high lying with a wide tween the houses indicating that they view to Voe of Browland to the East and were in function at the same time. South Loch of Grunnavoe to the South West, of house 17 there is a stone paving lead- and this could indicate that one or more ing up to the other standing stone. structures in fact are the remains of very dilapidated burial cairns. The water courses look as if they have been regulated, and West of the dyke

Nordlige Verdener Shetlandsprojekt 44

between house 1 and 2 there is a very wet area, and some of this dyke could have been damming a small lake at the time of the prehistoric use of Pinhoulland. Any kind of farming and especially the keeping of livestock demands a certain and steady amount of water (compare Edwards & Whittington 1998: 16; 17).

Dating Pinhoulland In the case of Pinhoulland, dating can only be verified through coring or excavation, but it is very likely that Pinhoul- land was in use during the Neolithic Pe- riod and during the Older Bronze Age. House 2, 6, 7, 9 and 11 belong to the pre-blanket peat strata, and thus must be older than Late Neolithic time or Old- er Bronze Age (compare Whittington 1980: 35). The many dykes connecting the house structures may indicate a relative chronology, as mentioned earlier. Preliminary analyses indicate at least a relative chronology in three phases. A scenario could be that House 6, 7 and 11 constitute phase 1 seen as single lying structures. It is impossible to say whether they could be contemporary other than 6 is younger than 11. is yet unknown, but what we see today is very probably a diachronic picture. Phase 2 could be the houses which are The cluster of houses (House 13, 14, 15, interrelated with dykes forming three 16, 17 and 18) are built together and groups of houses: House 1-2 as one; 3, with the exception of house 14 could 4, 19 and 5 as another interrelated group have been in contemporary use and con- and last House 9, 10, 12 as the third. Fig. stitute the third phase. The nearest par- 7 The chronology between these groups allels to the cluster of houses are 45 Photo Ditlev L. Mahler.

Fig. 5: House 12 seen from South East. The house mound with the Jarlshof house 8 and settlement types cluster of houses is seen behind house 12. such as Skara Brae from Orkney (Hamil- ton 1956: 9 ff.; Childe 1931: Plan of Vil- lage). Fig. 8 Jarlshof house 8 dated to Thus it is reasonable to presume at least the Older Bronze Age, and Skara Brae is as a model that Pinhoulland actually be- dated between c. 3100 and 2500 cal. BC gan some time during the Neolithic Pe- (Shepherd 2000:139; Clarke 1976: 26; riod, and considering the number of Ashmore 1998: 142-147; 2000: 299-308). houses, it could be quite early.

Nordlige Verdener Shetlandsprojekt 46 ± ±

1 15 14 1 16 18 15 14 13 ? 16 18 17 13 ? 12 17 12 10

10 7 9

7 9 Fig. 6: Pinhoulland phase 1? 11 Mapped by M. Hoydal/ 11 processed by P. Jensen. 2 6

2 6

19 8 5 19 3 5 8 0 25 50 100 4 3 Meters 0 25 50 100 4 Meters

Further considerations heather prevented a more thorough sur- During 2011 field work, many more sites vey. At Stanydale there are at least four were visited and surveyed but not nec- house structures which are likely to be- essarily mapped if the preservation con- ing older than the hall. One lonely struc- ditions were estimated to be poor or the ture not far to the West of the hall and vegetation was an obstacle as was the three house structures to the North of situation in the valley at Burwick, where the hall are bound together by dykes. 47 ±

1 15 14

16 18 1 13 ? 17 12

10 12

7 9 10 11 9

2 6

Fig. 7: Pinhoulland phase 2? 2 Mapped by M. Hoydal/ processed by P. Jensen. 19

5 8 3 19 0 25 50 100 4 5 8 3 Meters 0 25 50 100 4 Meters

There are also known house structures are interesting parallels to the suggest- to the South which will not be further ed relative chronology at Pinhoulland. mentioned here. The single house struc- At Skiords in Aithsting to the West of ture and the three bound together by Loch of Vaara on the way to Noonsburgh dykes – and two round field systems - we mapped a single house site and sev-

Nordlige Verdener Shetlandsprojekt 48

eral dykes among which was a smaller earlier observed. It is also suggestive circular dyke. The house is somewhat that the nearest parallel to the cluster of dilapidated with many boulders tumbled houses at Pinhoulland – as far as the in and the site could be of a somewhat surveys indicate – is Skara Brae and younger date. Further to the West at other of the clusters/villages of Late Ne- Ness of Nounsbrough and down to Gru- olithic or Older Bronze Age origins from ni Gill, by Ini Fiord and exposed to the Orkney. This also implies that we have South we mapped two house structures, to reconsider the cultural connections a circular dyke and other dykes indicat- between the North Atlantic islands, ing that the two houses could be con- which challenge the idea of isolation temporary. Fig. 9 during this period. We should all be well aware that all these indications are These few examples give a picture of a based on surveys and therefore are built variation in settlement structures: Sin- on peaty ground – so to say in “Shetlan- gle house structures lying alone; small dic” terms. communities consisting of two or more households with field systems; and finally Orkney has a rich and varied Neolithic clusters of houses such as the one at Period, which among other factors is Pinhoulland consisting of five to six struc caused by the intense interest from an- tures. As far as I know the house cluster tiquarians and archaeologists in Orkney at Pinhoulland stands alone on Shetland, prehistory over a considerable span of but further surveys and mapping will un- time. Shetland on the other hand has doubtedly reveal that this kind of settle- lacked the same magnetism seen from a ment type is more widespread. 19th and 20th Century view, which is no critique of the archaeological investiga- Conclusion tions in Orkney, but may be more an in- I have concentrated my work on West dication of where the prestige has been Mainland, Shetland and I am thus una- placed. The last 15-20 years has changed ble to say whether the above sketched much both concerning published exca- picture is representative for the settle- vations (Turner 1998; Fojut 2006; Dock- ment structure on Shetland as a whole rill et al. 1998: 61; AOC Archaeology during the Neolithic Period or Older Group), excavations especially of Iron Bronze Age. It looks as if Shetland has a Age – Viking Age structures and exhibi- much more complicated settlement struc tions among other things thanks to the ture during the Neolithic Period than investments in Hay’s Dock. ±

± 49

13 18 16

13 18 16 17

Fig. 8: Pinhoulland phase 3? Mapped by M. Hoydal/ processed by P. Jensen.

0 5 10 20

Meters

0 5 10 20

Meters Nordlige Verdener Shetlandsprojekt 50

Gruni Gill, Longaness, seen from the East. The site contains at least two house structures and a number of dykes among others a circular dyke as can be seen in the centre of the photo.

Photo Ditlev L. Mahler.

Acknowledgements Bibliography Alison Sheridan is thanked for the in- AOC Achaeology Group. 2007: The Neo- spiring discussions we have had both in lithic Heart of Shetland. Stage One Report. Edinburgh and on Shetland. Jenny Murray has been a great help during the field Ashmore , P. J. 1998: Radiocarbon dates work and in Lerwick Museum and to- for settlements, tombs and ceremonial gether with Laurie Goodlad and John sites with Grooved Ware in Scotland, Hunter we have had exciting excursions Prehistoric Ritual and Religion, Gibson, among others to sites we later decided A. & Simpson, D. (eds.), Stroud, Sutton mapping. Chris Dyer helped us with the Publishing: 139-147. mapping on Unst, and we received much support from Amenity Trust and Val Turner. Ashmore, P. J. 2000: Dating the Neolithic Finally I would like to thank Bjarne Grøn- in Orkney, Neolithic Orkney in its Euro- now for fruitful comments and Susan pean context, Ritchie, A. (ed.), McDon- Mahler for reading the proof. ald Institute Monographs, Cambridge: 299-308. 51

Bennett, K. D. et al. 1992: Holocene his- Edwards, K. J. & Whittington, G. 1988: tory of enviroment, vegetation and hu- Landscape and environment in prehis- man settlement on Catta Ness, Lunas- toric West Mainland, Shetland in Land- ting, Shetland. Journal of Ecology 80: scape History, vol 20: 5-17. 241-273. Edwards, K. J., Schofield, J. E., Whitting Calder, C.S.T. 1951: Report on the Exca- ton, G. and Melton, N. D. 2009: Palynol- vation of a Neolithic Temple at Stany- ogy ”On the Edge” and the Archaeologi- dale in the Parish of Sandsting, Shet- cal Vindication of a Mesolithic Presence? land. Proceedings of the Society of The case of Shetland, From Bann Flakes Antiquaries of Scotland (1949-50): 185- to Bushmills, papers in honour of Profes 205. sor Peter Woodman, Finlay, N., McCartan, S., Milner, N. and Wickham-Jones, C. (eds). Calder, C.S.T. 1956: Report on the dis- Prehistoric Society Research Paper No 1, covery of numerous Stone Age house- Oxford, Oxbow: 113-123. sites in Shetland, Proceedings of the So- ciety of Antiquaries of Scotland 89: Fojut, N. 2006. A Guide to Prehistoric and 340-97. Viking Shetland. Lerwick Times, Lerwick Gillmore, G. K. & Melton, N. 2011: Early Calder, C.S T. 1964: Cairns, Neolithic Neolithic Sands at West Voe, Shetland houses and burnt mounds in Shetland, Islands: Implications for human settle- Proceedings of the Society of Antiquar- ment. Geological Society, London, Spe- ies of Scotland 98 (1963-64): 37-86. cial Publications 352: 69-83.

Clarke, D. V. 1976: The Neolithic Village Grønnow, B., 2004: Saqqaqkulturen, at Skara Brae, Orkney: 1972-1973 Exca- Grønlands forhistorie, Gulløv, H. C. (ed.) vations. Edinburgh. Gyldendal, Copenhagen: 66-108.

Childe, V. G. 1931: Skara Brae, A Pictish Grønnow, B., ip: Independence I and Village in Orkney. London. Saqqaq: the first Greenlanders,Handbook of Arctic Arhaeology, Friesen, M. & Ma- Dockrill, S. J. Bond, J. M. and O’Conner, son, O. eds. Oxford: 1-23. T. P. 1998: Beyond the Burnt Mound: The South Nesting Palaeolandscape Project, Hamilton, J. R. C. 1956. Excavations at Shaping of Shetland, Turner, V. (ed.), Ler- Jarlshof, Shetland, Ministry of Works. wick: 61-82. Archaeological Reports no. 1. Edinburgh.

Nordlige Verdener Shetlandsprojekt 52

Kirch, P. V. 1988: Long-distance exchange land. Oxord University Comitee for Ar- and island colonization. The Lapita case. chaeology, monograph No 9, Oxford. Norwegian Archaeological Review 21, No. 2: 103-117. Shephed, A. 2000: Skara Brae: Express- ing Identity in a Neolithic Community, Mahler, D. L. 2011: Shetland: The Border Neolithic Orkney in its European Context, of Farming 4000-3000 BC. Features of the Ritchie A. (ed.), Cambridge: 139-158. Neolithic Period on Shetland, Farming on the edge: Cultural Landscapes of the Sheridan, A. 2010: The Neolithization of North. Some features of the Neolithic of Britain and Ireland: The “Big Piture”, Shetland. Short papers from the net- Landscapes in Transition, Finnlayson, B. & work meeting in Lerwick, Shetland Sep- Warren, G. (eds.), Oxford, Oxbow: 89-105. tember 7th-10th 2010. Northern Worlds. Mahler, D. L. & Andersen, C. (eds), Turner, V. 1998: Ancient Shetland. His- København: 6-18 toric Scotland. Shetland Amenity Trust. D. J. Breeze (ed.), London. Mahler, D. L. IP.: Shetland: The Border of Turner, V. 2011: From Homestead Enclo- Farming 4000-3000 B.C.E., Peopling an sure to Farm? Field Development in empty area? Papers from Nordic TAG Con Shetland in the Neolithic Periode. Farm- ference April 26th-29th 2010 in Kalmer. ing on the Edge: Cultural landscapes of the North. Short papers from the net- Melton, N. D. 2009: Shells, seals and work meeting in Lerwick, Shetland Sep- ceramics: an evaluation of a midden at tember 7th-10th 2010. Mahler, D. L. & An- West Voe, Sumburgh, Shetland, 2004-5, dersen C. (eds.), Copenhagen: 19-31. Mesolithic Horizons, Papers presented at the Seventh International Conference Whittington, G. 1980: A sub-peat dyke on the Mesolithic in Europe, Belfast in Shurton Hill, Mainland, Shetland. Pro- 2005, Vol 1 MaCartan, S., Schulting, R., ceedings of the Society of Antiquaries Warren, G. and Woodman, P., (eds.), Ox- of Scotland 109: 30-35. ford, Oxbow: 184-189. Whittle, A., Keith-Lucas, M., Miles, A., Rees, S. 1986: Stone Implements and Noddle, B., Rees, S. and Romans, J.C.C., Artefacts, in Whittle, A., Keith-Lucas, 1986: Scord of Brouster. An Early Agri- M., Miles, A., Noddle, B., Rees, S. and cultural Settlement on Shetland. Oxord Romans, J. C. C., Scord of Brouster. An University Comitee for Archaeology, Early Agricultyral Settlement on Shet- monograph No 9, Oxford. 53

Sacred Work: Cultivating the soil in prehistoric Shetland

Jenny Murray

The Shetland Isles are the most north- of wooden tools which date to the Iron ern archipelago in Britain; an exposed Age, suggesting that early farmers could situation where the North Sea meets have had access to organic implements the Atlantic, offering little relief from which may not have survived in the ar- the numerous gales that blasts its chaeological record. shores. Yet, people have farmed this small group of islands for thousands of Today the Shetland landscape offers years. The landscape is peppered with limited areas of fertile ground. Arable small clearance cairns and field bounda- pockets of land lie close to the shoreline ries; a testament to the Neolithic popu- and palynological studies suggest some lation who first delved its soil, clearing of these green patches have been areas to plant and nurture a life sustain- farmed since the Neolithic period (Whit- ing crop. This paper will explore the tle et al. 1986, Owen & Lowe 1999). The tools these early farmers may have used earliest signs of anthropogenic impact to break the ground and till the soil. It on the landscape include deposits of will also explore the possible symbolic charcoal underneath a prehistoric stone significance of these tools to the people dyke on Shurton Hill near Lerwick. Ra- who used them. diocarbon dated to c. 2800 BC (UB2122), this evidence of deliberate burning may The Shetland museum has a vast collec- suggest heathland clearance in an at- tion of stone tools which have survived tempt to maintain grazing pasture (Whit- from various prehistoric sites around the tington 1977: 33). Early farmers at the isles. Many of these may have been Scord of Brouster were also clearing used to delve the land, but recent re- land for agriculture by burning, around search has revealed a small collection 3000-2500 BC (Whittle et al. 1986:

Nordlige Verdener Shetlandsprojekt 54 Photograph by the author. Fig. 1: Stone ard share from Scord of Brouster.

108,146). Cultivation by ard at this site at Sumburgh, Fair Isle and Kebister (Lamb is evident from five broken stone ard & Rees, 1981: 117, Hunter 1996: 51, Owen points recovered from the construction & Lowe 1999: 281). of the earliest dwelling (House 2) and charred barley, radiocarbon dated to Stone tools 2590 ± 65 BC (CAR-251), (Whittle et al While evidence for ploughing by ard is 1986: 112). A total of 75 ard points were reliable on Neolithic sites such as Scord recovered from Scord of Brouster, last of Brouster, tools used to break the fal- occupied around 1400-1300 BC (Rees low land are not so obvious in the ar- 1986: 74, Whittle et al. 1986: 34-37). chaeological record. There are a range of stone tools that could have been used Further emphatic evidence for early cul- to manually break the sod including a tivation includes ard marks or plough selection of ‘spades’ on display in the furrows discovered in the subsoil during Shetland Museum (see fig. 2). This in- excavations of Bronze and Iron Age sites cludes a heart shaped tool with a hole at the top which may have been hand- Photograph by Dr. Simon Clarke. held. Several of these were found in the 1930s by A. O. Curle who was excavat- ing at the multi-period site at Jarlshof in Fig. 2: Ard marks Sumburgh. He concluded their shape clearly visible in the ancient soils at was reminiscent of a labourer’s shovel Scatness, Shetland. which suggested they may have been These marks predate used to work the soil (Curle 1933: 101). the settlements de- fensive ditch which However, he found there was little evi- cuts through them. dence of wear around the edge to sug- 55 Photographs by the author. Fig 4: Heart shaped ‘shovel’ (ARC 65323).There is little evidence on the cutting edge to suggest it was used for this purpose.

Fig. 3 (left): A selection of possible rudimentary ‘spades’ on display in the Shetland Museum. gest this employment; in fact the only tary mattock or hoe (2006: 30). She fur- wear appeared to be at the handle edge ther proposes that flaked stone bars and (as evident in fig. 4). These tools were ards should be considered together as looked at again by lithic specialist Ann tools for cultivation as both are com- Clarke, who suggested they may not have monly found placed together as inclu- been shovels but rather a veggewol, a stone object that was built into the walls Fig. 5: Flaked stone bars from House 2, Scord of Brouster. of the byre to secure an animal, this would explain the wear at the handle and not the cutting edge. (2006: 38)

Another group of tools which may have been more suitable for breaking ground are the flakes stone bars, of which there are many hundreds in the museum collection. Ann Clarke’s research showed that many of these presented signs of being hafted, and wear was noted at one side of their cutting edge which may suggest these were used as a rudimen- Photograph by the author.

Nordlige Verdener Shetlandsprojekt 56

sions into stone structures at Neolithic Wooden and bone tools and Bronze Age settlement sites. They From the Neolithic period, as noted appear together at the Scord of Brouster above, tilling of the soil was done by as early as 3500 BC (2006: 111,130). ard, but recent evidence also suggests This practice may allude to the impor- that, by the Bronze Age, some turning of tant social value of these agricultural the soil was done by other tools. Spades tools to the people who relied on them. are almost absent from the archaeologi-

Fig. 6: cal record but inferential evidence for Hand-held The profile of other stone tools in the their use is observed in ancient field sys- tools from Scord of Brouster assemblage lend tems. Grooves in the Bronze Age subsoil Scord of Brouster and themselves to be hand-held and may during excavations at Cornwall, in south- Sandsting have been practical for breaking up ern Britain, suggest the soil appears to (ARC 8056). clods on very small plots (see fig. 6). This have been broken by a spade or hoe Could these have been would have been arduous work, but rather than the ard (Rees 1979: 329; used to break work was most likely done in groups. Lerche & Steenberg 1980: 63). Similar up clods? ridges and furrows dating to 1500 BC have been uncovered in areas of blanket- bog in Western Ireland (Mitchell 1978: 27). Likewise, similar evidence was noted in Iron Age field systems in Denmark (Lerche & Steenberg 1980: 63). In Scot- land, patches of cord rig, that resemble miniature rig-and-furrow, have been revealed in Iron Age field systems; these appear to have been hand dug (Armit & Ralston 2005: 190).

Further north in Shetland, excavations of a Bronze Age house at Mavis Grind revealed features similar to spade-dug Photographs by the author. lazy beds, dug through several centime- tres of accumulating peat (Cracknell & Smith 1985: 91). Also, recent excavations at Old Scatness revealed marks in the sub-soil which were shorter and less sharp in profile than ard furrows – these 57 Photograph by Lizzie Ratter. Fig. 7: Shetland spade from the Late Iron Age period. Photograph by the author. Fig. 8: Iron Age clod breaker found buried with the wooden spade in Fig. 7 (ARC 65122).

suggest spade use in an Early/Middle carbon dating showed these tools to be Iron Age context (Dockrill et al. forth- much earlier than previously anticipated. coming). A sample from the Yell spade was sent to the laboratory for dating and returned Within the museum collection are two a date from Shetland’s Late Iron Age pe- wooden foot-spades and a clod breaker, riod – AD 570±30 (GU23299). The Unst uncovered from deep moor by people clod breaker was dated a little earlier, cutting peats on the hillside (see figs. 7 240-400 cal AD (GU23300). and 8). One of the spades, from the island of Unst, was buried with the clod Evidence from the field, plus the recent breaker suggesting these tools were used dating of these tools, may suggest that together. Another spade was buried on wooden spades could have been used in the nearby island of Yell. They are all in Shetland since the Bronze Age period. It excellent condition having been perfect- has been suggested that the spade is ly preserved in the anaerobic conditions more effective than the ard for tilling which peat moor offers. Recent radio- the soil, in terms of output per acre

Nordlige Verdener Shetlandsprojekt 58

(Steensberg 1986: 110-111). This meth- branches. Unfortunately these are ab- od would certainly be appropriate for sent from the archaeological record in smaller areas of arable ground. While places such as Scord of Brouster; they turning the soil you could pulverise may have not survived, perhaps ending lumps and turn in manure, providing the their useful life as fire wood. perfect tilth for crops to grow (ibid). Again, teamwork would have been ap- Other organic implements, such as bone plied making the task easier and less tools, may have also been lost to the boring as shown in recent times when record due to the acidic soil conditions groups of able Shetlanders gathered to in certain areas. Excavation of the earli- delve their fields by spade (see fig. 9) est layers at Jarlshof revealed two shovels crafted from the shoulder blades of oxen (Hamilton 1956: 12). Unfortunately, excavations at Scord of Brouster produced a very small assemblage of bone due to the acidic conditions of the site (Noddle 1986: 132). Despite the small sample analysed, the majority appeared to be cattle (ibid). It is therefore possible that bone shovels may well have been used at this settlement to break Fig. 9: In modern times, tilling the land was tradi- the fallow, but have not survived. Anna tionally done by the Shetland spade as seen here Ritchie’s excavations at the Neolithic at Springfield in Fair Isle, taken circa 1920. Image courtesy of Shetland Museum and Archives. site at Knap of Howar, on Papa Westray, in Orkney revealed a possible whale- bone hoe (Ritchie 1983: 52), and these While the evidence for wooden spade examples demonstrate that a variety of use does not date back to Shetland’s faunal material may have been utilised Neolithic period, other wooden imple- in the construction of cultivation tools. ments such as digging sticks may have been utilised, similar to the example The examples explored above highlight discovered in a bog at Døstrup, Den- three very different resources used in the mark, dating to cal.810 BC (K-3266), manufacture of tilling implements. Stone (Lerche 1995: 180). These simple but ef- examples have survived in their hun- fective tools could have been manufac- dreds; while many others, such as bone tured in Shetland using small trees and and wood, may have been lost in time. 59

The significance of agrarian man fertility, become more evident from tools in prehistoric Shetland the Bronze Age and continue into the Many of the tools discussed have been Iron Age period (Bradley 1998: 183). uncovered, not in fields and cultivated These offerings, such as the wooden areas where you would expect to find tools, were often placed into wetland them, but from the walls of structures in localities, often bogs, streams, rivers or settlement areas and further afield and in lakes. Were these perhaps an offering deep moorland. They have been carried to higher powers to ensure future suc- back from the fields by early farmers and cessful harvests from the surrounding deliberately placed there. The reasons fields? Two millennia later, these bog- for this may be quite significant. lands offered Shetlanders a valuable resource for peat cutting, and the eventual Ritual deposition of tools and everyday uncovering of these votive deposits objects is well known throughout pre- which lay hidden in their depths. history. When C.S.T. Calder excavated Neolithic period remains on the west These physical components of everyday side of the Shetland mainland during life appear to have taken on a symbolic the 1950s, he found the house walls role as metaphors for regeneration, not were simply littered with rude stone im- only for the land but also the fertility of plements; 600 were recovered from the the community. Archaeologist Richard Ness of Gruting and a further 250 from Bradley suggests these offerings asso- Stanydale (Calder 1955: 356). Similarly, ciated with agricultural produce laid a clearance cairns from the surrounding special significance on the land as well field systems in all the areas looked at as the people (1998: 171,183). These, he by Calder also contained deposits of states are a series of rituals that stressed stone tools; one local man who demol- the importance of food production, es- ished a clearance cairn at Lower Gruniq- pecially in Iron Age society, and the dep- uoy collected 26 rude implements (Cal- osition of agricultural equipment en- der 1955: 357). Ann Clarke suggests the sured the fertility of the crops was inclusion of ard points and flaked stone protected (Bradley 1998: 170-171). bars into structures afforded them a deeper meaning, ‘as a symbol of place, Conclusion or of the past’ (2006: 125). Food production forms the very basis of a settled society, and from field systems of Ritual offerings with agricultural conno- prehistoric Shetland we have, in the local tations, linked with cultivation and hu- museum, a large selection of tools used

Nordlige Verdener Shetlandsprojekt 60

to cultivate the land by early farmers. Curle, A. 1933: Account of Further Exca- These include the large assemblages of vations in 1932 of the Prehistoric Town- rough stone tools; skillfully produced ard ship at Jarlshof, Shetland, on behalf of shares; finely crafted wooden spades; H.M. Office of Works, Proceedings of the clod breaker and rare bone tools. the Society of Antiquaries Scotland Vol. These have all survived because of their 67 (1933): 86-136. significance to the community who used them, by building them into the walls of Dockrill, S., Bond, J., Turner, V., Brown, their dwellings and field systems, and L. et al. (forthcoming). Excavations at by offering them to higher powers in the Old Scatness, Shetland Volume 2: The surrounding landscape. The study of Iron Age Broch and Village. these not only offers us an insight into the cultivation tools used in Shetland’s Downes, J. & Lamb, R. 2000: Prehistoric prehistory, but we can now gain a deep- Houses at Sumburgh in Shetland, Exca- er understanding of the lives and beliefs vations at Sumburgh Airport 1967-1974. of those who last touched them. Oxford.

Hamilton, J. R. C. 1956: Excavations at Bibliography Jarlshof, Shetland. Edinburgh.

Bradley, R. 1998: The Passage of Arms, Hunter, J. R. 1996: Fair Isle, The Archaeo an Archaeological analysis of prehistor- logy of an Island Community. Edinburgh. ic hoard and votive deposits. Oxford. Lerche, G. 1995: Radiocarbon Datings of Calder, C. 1955: Report on the Discovery Agricultural Implements, Tools and Till- of Numerous Stone Age House-sites in age 1968-1995, Revised Calibrations Shetland. Proceedings of the Society of and Recent Additions. Lerche, G., Fen- Antiquaries Scotland, Vol. 89 (1955- ton, A., Steensberg, A. 1975. Tools and 1956): 340-397. Tillage Vol. VII:4, Copenhagen, National Museum of Denmark. Clarke, A. 2006: Stone Tools and the prehistory of the Northern Isles. Oxford. Mitchell, F. 1978: Tillage Tools in Ireland during the Past 5000 years. Journal of Cracknell, S. & Smith, B. 1985: Excava- Earth Sciences 2(1): 27-40. tion at Mavis Grind. Smith, B. (ed.), Shet- land Archaeology. Lerwick: 85-95. 61

Lamb, R. & Rees, S. 1981: Ard Cultiva- Rees, S. 1986: Stone Implements and tion at Sumburgh, Shetland. Lerche, G., Artefacts. (Referred in:) Whittle, A., Keith- Fenton, A., Steensberg, A. Tools and Till- Lucas, M., Milles, A., Noddle, B., Rees, S., age Vol. IV:2, Copenhagen: National Romans, J., Scord of Brouster, An Early Museum of Denmark: 117-121. Agricultural Settlement on Shetland, Ex- cavations 1977-1979. London: 75-91. Lerche, G. & Steensberg, A. 1980: Agri- cultural Tools and Field Shapes. Copen- Ritchie, A. 1983: Excavations of a Neo- hagen: National Museum of Denmark. lithic farmstead at Knap of Howar, Papa Westray, Orkney. Proceedings of the So- Noddle, B. 1986: Animal Bones. (Re- ciety of Antiquaries Scotland, Vol. 113: ferred in:) Whittle, A., Keith-Lucas, M., 40-121. Milles, A., Noddle, B., Rees, S., Romans, J., Scord of Brouster, An Early Agricul- Steensberg, A. 1986: Man The Manipu- tural Settlement on Shetland, Excava- lator. Copenhagen, National Museum of tions 1977-1979. London: 75-91. Denmark.

Owen, O. & Lowe, C. 1999: Kebister: The Whittle, A., Keith-Lucas, M., Milles, A., four-thousand-year-old story of one Shet Noddle, B., Rees, S., Romans, J. 1986: land Township. Society of Antiquaries of Scord of Brouster, An Early Agricultural Scotland Monograph Series Number 14. Settlement on Shetland, Excavations Edingbourgh. 1977-1979. London.

Rees, S. 1979: Agricultural Implements Whittington, G. 1977: A Sub-peat Dyke of prehistoric and Roman Britain. British on Shurton Hill, Mainland, Shetland. Archaeology Reports (Series 69), Vol, 2. Proceedings of the Society of Antiquaries Oxford. Scotland Vol. 109: 30-35.

Nordlige Verdener Shetlandsprojekt 62

The distribution of worked felsite – within and outwith Neolithic Shetland

Torben Bjarke Ballin

Introduction that Orkney formed an integral part of In a previous paper (Ballin 2011d), the an extensive exchange network. author discussed the dating and character of the initial colonization of Shetland. One of the key artefact groups discussed This discussion was based on the pres- in Ballin (2011d) was felsite implements. ence and absence of diagnostic material They almost exclusively include axe- culture elements, with the evidence heads and Shetland knives, with some suggesting rather late colonization, pos- scrapers, and possibly arrowheads, hav- sibly around the Mesolithic/Neolithic ing been made on the basis of fragments transition, and with the bulk of the earli- of recycled polished axeheads. The aim est evidence dating to the later part of of the present paper is to gain more in- the Early Neolithic period. The almost sight into the dynamics of Shetland’s complete absence of Shetland artefacts Neolithic society by analyzing the pro- south of the island group, in conjunction curement and distribution patterns of with the almost complete absence of ar- felsite – from the quarries in North Roe tefacts from southern Scotland in Shet- to the end-users, within as well as out- land, indicates that Neolithic Shetland with the Shetland Isles. may have been in a state of complete or partial isolation, in contrast to for exam- The organization of the ple Neolithic Orkney which boasts nu- North Roe mining operations merous exotic objects, such as artefacts When the physical features of prehistor- in Arran pitchstone and Yorkshire flint ic quarries are characterized, analysts (Ballin 2009; 2011c), clearly showing tend to follow classification schemas 63

Fig. 1: Axehead production at the Beorgs, showing crude waste, like that of Schneiderman-Fox & Pappa- as well as one of several oval shelters. Courtesy of Gabriel Cooney, University of Dublin. lardo (1996), emphasizing the following activity areas:

• the quarry itself where material is • the lithic reduction site above the extracted; quarry face or on a level terrace ad- • the tailing pile, just below the quarry jacent to the quarry face, where re- face, containing blocks of quarried duced blocks are further reduced material; into preforms or final tools. • the ore dressing, milling, or transition area, located below and within In his various papers on the North Roe 50 m of the quarry face, where large quarries (Ballin 2011a; 2011b), the author blocks are broken down for trans- has tended to characterize much of the port; and area’s felsite debris as forming tailing piles, supplemented by a number of ob-

Nordlige Verdener Shetlandsprojekt 64

vious workshops. This, however, is not an Immediately adjacent to the felsite dykes, entirely acurate depiction of the North workshops were organised to allow fur- Roe situation. In many cases, the quarry ther reduction of selected blocks of raw waste does appear to form down-hill material. screes – that is, traditional tailing piles – but in most cases, the linear character Detailed surveys of the main North Roe of the debris is a result of the size and procurement centres – the Beorgs of shape of the outcrops, mostly occurring Uyea ridge in the north and the Midfield as kilometre-long dykes, and not gravity. hill in the south – showed that some Basically, it seems, dykes were mined, and specialization may have taken place in quarry waste deposited either behind the prehistory. The former centre appears to on-going quarrying operation – that is, have focused predominantly on the pro- on top of already emptied sections of the duction of axeheads, and the latter on dykes – or next to the exhausted dykes, the production of Shetland knives. Indi- resulting in linear accumulations of waste. cators of this are many, such as axehead

Fig. 2: Denticulated boulder on Midfield – probably an abandoned waste ‘core’ from the production of knife tool blanks. Cour- tesy of Gabri- el Cooney, University of Dublin. 65

Fig. 3: Shetland knife workshop on Midfield, with knife roughout and production waste. Courtesy of Ditlev L. Mahler, The National Museum of Denmark. and knife preforms and different forms represent waste and the core becomes of waste. The waste relating to axehead the intended implement, whereas the production is generally somewhat larger latter is based on the procurement of and cruder than that relating to the pro- flake tool blanks, with the resulting core duction of knives, fig. 1. being the waste product.

Although the construction of truly relia- Fig. 1 shows debris from axehead pro- ble operational schemas for the two dif- duction (blocks, large misshapen flakes, ferent types of production will have to and bro-ken or rejected axehead rough- await the excavation and analysis of dif- outs), whereas figs. 2 and 3 show waste ferent (axehead and Shetland knife) from the production of Shetland knives. workshops, there are indications that the The denticulated boulder in fig. 2 may two manufacturing forms may have fol- represent the manufacture of flake lowed fundamentally different schemas. blanks for knives, where large flakes Most likely, the production of felsite axe were detached from the edge of this im- head was generally carried out in the form pressive parent piece (technically, an of ‘core production’ and the production of over-sized ‘core’), whereas Fig. 3 shows felsite knives as ‘flake production’. The small-sized production waste surround- former is based on the removal of flakes ing a Shetland knife rough-out, immedi- from a block of rock, where the flakes ately left of centre. Fig. 2 & 3

Nordlige Verdener Shetlandsprojekt 66

Fig. 4: Almost finished (not ground or polished) felsite axehead. Courtesy of Ian Tait/Jenny Mur- There may be many reasons for this sep- ray, Shetland Museum. aration of processes, either practical and/or ideological, but it is almost cer- However, operational specialization did tain that the procurement of felsite in not only take place within the North Roe North Roe included some religious con- quarrying complex. It is also obvious siderations, and that the quarrying of that differential specialization charac- felsite was, at least to a degree, ritual- terized the relationship between the ized. Pete Topping’s (2005) analysis of North Roe area and domestic settle- flint quarrying at Grimes Graves in Suf- ments outside this area. Till this day, no folk suggests ritualization of flint pro- pol-ished felsite axeheads or Shetland curement at this extensive location, and knives have been recovered from North he presents the following generalized Roe, with all polished felsite objects ‘cycle of events’: 1) cleansing rituals, 2) having been recovered from sites, or as offerings, 3) extraction, 4) post-extrac- stray finds, south of Ronas Voe. This tion prayers and offerings, 5) artefact suggests that, in North Roe, blocks of production, 6) ceremonial use of [some] felsite were mined and axehead and artefacts, and 7) rites of renewal. knife rough-outs produced, but that these preforms were ground and pol- Scott & Thiessen’s (2005) analysis of ished (that is, made into final tools) out- catlinite extraction (for ceremonial pipes) side the quarrying complex. Fig. 4 at a location in Minnesota puts forward 67

a list of rituals to be carried out in con- more felsite in some parts of the Island nection with the pro-curement of this group than in others. However, it is clear raw material. Although it should be re- that felsite occurs throughout Shetland in membered that catlinite was intended a number of different contexts, namely exclusively for use in the ceremonial 1) as part of assemblages from domestic sphere, their account of the social con- settlements, 2) as burial goods, 3) in the text of the quarrying operations is in line form of caches, and 4) as stray finds. with the social contexts defined in relation to the mining of ‘general purpose’ At domestic settlements, felsite is usu- raw materials (such as Grimes Graves ally either absent, or it occurs in small flint; Topping, above). Consequently, numbers. The settlement finds usually their list of ‘attendant rituals’ may have include fragments of felsite, a few axe- implications for the understanding of head or knife fragments, or a limited prehistoric raw material procurement in number of smaller tools, frequently general: 1) the quarry is sacred and the based on parts of cannibalized felsite quarriers therefore camp away from the axeheads (eg, at Scord of Brouster in outcrop; 2) a three-day purification ritual West Mainland, and at Scatness Brough is carried out; 3) offerings are made to in Sumburgh; Ballin 2005 [Scord]; 2008 propitiate the guardian spirits and seek [Scat rep]). The cist burial at Sumburgh permission to extract the wanted stone; (Hedges et al. 1980) includes one small 4) sexual relations with women are to axehead which has been classified as be avoided during this period; 5) exclu- felsite; however, this classification is sion of women from the quarry while based on simple visual inspection and mining was on-going; and 6) the quarrier the piece should be either thin-sectioned had to be a man above reproach (com- or analysed by XRF. The most well- pare this list with Topping’s‘cycle of known felsite deposition is arguably the events’, above). cache from Stourbrough Hill in West Mainland (Fojut forthcoming), where The Distribution of felsite across Noel Fojut came upon 19 finely polished Shetland – further modification, unused Shetland knives, stacked verti- use and recycling cally like books on a shelf. It is uncertain So far, no maps have been produced to whether this is a ‘trade’ cache or a ritual show the distribution of worked felsite deposition. However, the most numer- not to mention: the different classes of ous group of felsite objects are the stray felsite implements across Shetland, and it finds, which include many fragments, as is presently uncertain whether there are well as intact implements.

Nordlige Verdener Shetlandsprojekt 68

The Shetland axeheads (almost all stray The latter may be an over-generaliza- finds) were examined and summarily tion, as some of the smaller felsite axe- characterized by Ritchie (1992). He notes heads are obviously functional imple- that the Shetland axeheads are gener- ments. Particularly small adzes seem to be ally considerably larger (in several cases functional pieces, being not only smaller with lengths of 200-300 mm) than the and less elegantly shaped/finished, but axeheads from, for example, Orkney, and also occasionally with marks from use. that felsite axeheads are generally con- An assemblage from Lagan Tormore at siderably larger than pieces in Cumbrian Sullom Voe (presently being processed; tuff or Antrim porcellanite. They are also Ballin forthcoming [LT rep]) includes two usually neatly finished. He states that intact adzes, as well as several other ‘… the purpose of the Shetland axe- felsite implements and various froms of heads in general poses questions. Such felsite waste (some probably from ad- information as is presently available justing the adzes’ butts for haft-ing). The shows that [Shetland was] typified by author characterized the two objects in birch-hazel scrub, and one wonders why the following way (ibid.): early Shet-landers needed such large axeheads. Apart from being large, many ‘In functional terms, both pieces are of the Shetland axeheads are very well adzes (L = 118 mm and 136 mm), and made, sometimes with splayed cutting they were hafted with their working- edges. They show little or no sign of use edge at a perpendicular angle to the and give the impression of being cere- axe-shaft. This is indicated by their monial or prestige pieces’. cross-section (plano-convex), edge-shape

Fig. 5; Adze from the stores of Shetland Museum – the axeheads from Laggan Tor- more belong to this formal cate gory. Courtesy of Ian Tait, Shetland Museum. 69

(hollow-edged), profile category (asym- To fully understand the production, dis- metrical), and working-edge profile (asym- tribution and use of felsite artefacts metrical). Fig. 5 Both axeheads are fully within Shetland, it is essential that, in polished, but they also differ on several the future, felsite axeheads and knives points: where one (CAT 1078) has a are subdivided into formal classes some splayed edge, the other piece (CAT 1079) possibly being functional and some cer- has parallel lateral sides, and where the emonial, and that distribution maps are former has a rounded/pointed butt, the produced, showing the exact dissemina- latter has a flat/flat butt – that is, flat tion of these pieces across the island seen from above and flat seen from the group. side. CAT 1078 needed minimal additional modification to allow it to be fit- The distribution of felsite outwith ted into an existing axe-shaft, whereas Shetland – comparison of the CAT 1079 was substantially altered: to felsite exchange net-work with allow hafting, its butt was made consid- other contemporary networks erably narrower by the detachment of As mentioned above, felsite practically several relatively large flakes by strikes never left Shetland (no knives and very to its lateral sides. This probably indi- few axe-heads), and an effort was ap- cates that 1) these axeheads were work- parently made to retain artefacts in this ing-implements, and not pieces for ritu- raw material within the Shetland social al deposition, and 2) on Shetland with territory. Below, a number of other raw its sparse growth of trees, good wooden materials and their distribution pat-terns shafts were more valuable than the pol- are investigated to find out whether this ished working-axeheads’. practice was common in Neolithic times, or whether it sets Shetland apart as in As mentioned above, it is thought that some way following a different – more Shetland axeheads and knives in felsite isolationist – path to paths chosen by were polished outside the main procure- other social groups in Britain and North- ment area in North Roe, but at present it west Europe. is uncertain whether this happened at specialized sites - for example in coast- The main factor influencing the ex- al parts of North Roe – or at ordinary change of raw materials in the Neolithic settlement sites. So far, there is no evi- period was probably whether the indivi dence to support large-scale polishing dual raw materials were associated pre- of felsite artefacts at any of the known dominantly with functional or symbolic domestic sites from Shetland. values. Exchange of raw materials asso-

Nordlige Verdener Shetlandsprojekt 70

ciated with functional values tends to fracture as well as numerous cracks, be characterized by gradually declining and it is generally perceived as a rela- fall-off curves, whereas exchange of tively poor raw material (Kindgren 1991; raw materials asso-ciated with symbol- Högberg & Olausson 2007: 132). It is ic values tend to be characterized by found at the Kinnekullen peak, near the more complex fall-of curves (cf. Ballin Hornborga Lake, from where it was dis- 2009 [pitch]). tributed throughout a zone with a c. 100 km radius, Fig. 6. The fall-off curve for The distribution of Cambrian (or Kinne this resource, whether linear or logarith- kullen) flint in central Sweden is an emi mic, is gradually as well as steeply de- nent example of the first form of regres- clining, and both curves are associated sion. This type of flint, which was used with a high correlation coefficient, show- and exchanged in the Late Mesolithic ing an almost perfect relationship be- Lihult period of western and central tween distance and quantity (R2 = 0.91 Sweden, is characterized by conchoidal and 0.86, respectively). It is clear that, in this case, a particular low-grade raw 100 material was ex-ploited heavily near the

90 source but, with growing distance to source, it was gradually replaced by oth- 80 er raw materials which were obtainable at a lower price if obtained by exchange 70 or input of labour if obtained by direct procurement. The exaggerated steep- 60 y = -53.84ln(x) + 244.83 ness of the curve is an indicator of the

50 R² = 0.8588 relatively low value of this resource. Per cent 40 Chert another somewhat flawed raw y = -1.3906x + 112 material from southern Scotland (Ballin & 30 R² = 0.9096 Johnson 2005) is characterized by a similar form of fall-off curve, where settle- 20

10 Fig. 6: Linear and logarithmic fall-off curves for 0 Cambrian flint from the Kinnekullen peak in central Sweden (based on information from Kindgren 0 20 40 60 80 100 1991, Table 4). The correlation coefficients of the Distance two curves have been calculated. 71

30 Relative distance to quarry 25

TN - far from Bømlo 15

TN - near Bømlo 5

0 10 20 30 40 50 60 70 80 Rhyolite ratio Fig. 7: The distribution of worked rhyolite in relation to the Bømlo quarry complex.

ments in the interior of this region are Neolithic), Arran pitchstone (mainly Early dominated completely by the use of the Neolithic), and Yorkshire flint (mainly Late locally abundant chert, whereas the pro- Neolithic). portion of this raw material declines gradually and rapidly towards the coastal zone In Southwest Norway, Bømlo rhyolite where pebble flint with considerably was introduced quite rapidly at the be- better flaking properties and offering more ginning of the Early Neolithic, and it was durable cutting edges was available. used for blade and small-tool production throughout this period. Fig. 7 shows the With an abrupt stop to the exchange of distribution and ratio (rhyolite propor- felsite at the coastal borders of the tion of assemblage) of worked rhyolite Shetland social territory, it is almost cer- in relation to the Bømlo quarry complex, tain that felsite was considered more and it is possible to note several trends than simply a functional raw material, – although the relatively low number of and it is therefore most appropriately sites should be borne in mind: 1) Near compared to exchanged Neolithic raw the quarry complex, rhyolite makes up c. materials with similar non-linear fall-off 50-80% of assemblages; 2) further from curves, such as Bømlo rhyolite (Early the quarries, rhyolite makes up c. 20-

Nordlige Verdener Shetlandsprojekt 72

600

500

400

300 Number

200

100

1-10 21-30 41-50 61-70 81-90 101-110 121-130 141-150 161-170 181-190 201-210 221-230 241-250 261-270

Distance to source Fig. 8: The distribution of pitchstone outwith the central area (Arran/Argyll and Bute) of the exchange network.

50% of assemblages; 3) at a certain dis- macro-territory, which was sub-divided tance, the rhyolite ratio drops abruptly to into several social micro-territories pos- 0%. Fig. 7 If addition of further evidence sibly representing the ‘tribe’ and either supports this pattern, the following could ‘clans’ or ‘lineages’, respectively. be suggested: 1) The distribution of rhyo lite is most certainly not linear (in con- The abrupt drop in the rhyolite ratio ap- trast to the distribution of Cambrian parently coincides with natural markers in flint, above), and the relatively high rhyo the landscape, such as deep or broad fiords. lite ratio throughout Southwest Norway, Although the topography of Southwest Nor- in conjunction with the abrupt drop in way and Shetland differ somewhat, the ratio at a certain distance, indicates that two distribution patterns have obvious rhyolite may have had an emblematic similarities, like the abrupt drop in fre- function, that is, as a material marker of quency at a potential territorial border, with group identification; 2) the two-step dis- near-zero distribution outwith this area. tribution pattern in Fig. 7 may indicate that Pitchstone, a raw material for small-tool Southwest Norway represents a social production, was used on the source is- Cambrian flint Southern Scottish chert

Period Mesolithic Mesolithic/Early Neolithic

Used for Small tools Small tools

Extent of exchange network The Kinnekullen area (radius c. 100km) Central and southern Scotland

Distribution pattern Steep linear regression when alternative Steep linear regression when alternative raw materials become available raw materials become available

Perceived as Largely functional Largely functional

North Roe felsite Bømlo rhyolite

Period Later Neolithic Early Neolithic

Used for Axeheads and Shetland knives Small tools

Extent of exchange network Shetland Islands SW Norway

Distribution pattern Shetland only SW Norway only

Perceived as Emblematic (social ID) Emblematic (social ID)

Arran pitchstone Yorkshire flint

Period Early Neolithic (off Arran) Late Neolithic

Used for Small tools Axeheads and small tools

Extent of exchange network Northern Britain (less Shetland) Northern Britain (less Shetland)

Distribution pattern Multi-peaked fall-off curve Excl. use in S / centralized distr. in N

Perceived as Emblematic / prestige-related exotica Functional / prestige-related exotica

Table 1. Brief summaries of the exchange networks discussed in this section.

land of Arran possibly including Argyll & mainlands, redistribution occurred via Bute throughout prehistory, Fig. 8. The very large centres, supplying extensive exchange in pitchstone outside this area areas of hinterland. The frequency of is mainly an Early Neolithic phenomenon, pitchstone clearly declines with growing and it is characterized by several peaks. distance to the sources on Arran, and it The distribution is extensive, with finds is possible to suggest a zonation of Scot- of pitchstone having been made as far to land / northern Britain based on this fact: the north as Orkney and as far south as Arran itself represents one zone, char- the Isle of Man and southern Cumbria acterised by very high proportions of (Ballin 2009). pitchstone and use of volcanic glass throughout the Mesolithic, Neolithic and It is thought that the exchange of Arran Early Bronze Age periods; a zone around pitchstone may have been organized in Arran – involving the western half of a complex network based on redistribu- southern Scotland and Northern Ireland – tion centres. The fall-off curve in Fig. 7 is characterised by the presence of vast suggests that, on the Irish and British centres, each counting more than 500

Nordlige Verdener Shetlandsprojekt 74

pieces within one 10 x 10 km square from the Overhowden Henge in the Scot- (Fig. 8); in a third zone – Southeast Scot- tish Borders. This raw material was used land and the area around the Firth of during the Late Neolithic period for the Forth near Edinburgh – pitchstone is still production of small-tools as well as flint relatively common, but it does not occur axeheads, and, towards the north, the ex in these exceptional numbers; and in a change network extends from the greater peripheral zone up to 400 km from Arran, Yorkshire area to Orkney as in the pitch pitchstone-bearing sites are character- stone case, Shetland seems to have been ised by the presence of, at most, one or excluded from the exchange network. two pieces apart from on Orkney, where pitchstone artefacts are relatively nu- It is difficult to fully comprehend the merous in assemblages from ritual and meaning of the Yorkshire flint exchange, high status sites. Table 1 as the Late Neolithic sites investigated in Scotland may not be representative This distribution pattern differs consid- of the full range of site types: apparent- erably from those of the rhyolite and fel- ly, many known locations are either ritu- site exchange networks. Pitchstone may al sites or high-prestige settlements have had an emblematic function to the rather than traditional domestic sites. people (tribe) controlling the raw mate- Presently, finds from southern and north- rial sources (with Arran/Bute being the ern Scotland suggest that Yorkshire flint only part of Scotland where assemblag- may have been perceived differently at es may be exclusively in, or dominated the centre, and near the periphery, of by, pitchstone, possibly identifying peo- this vast exchange network: in southern ple within this social territory as ‘those Scotland – that is, the Scottish Borders, who use pitchstone’), but in the bigger South Lanarkshire, and the Lothian coun- picture pitchstone exchange may have ties – lithic assemblages are almost ex- been the ‘glue’ by which numerous so- clusively in this material, indicating that cial territories were held together in a it may have been perceived predomi- single over-arching social structure or nantly as a functional mate-rial, where- contact network. Within Scotland, only as in the north – for example on Orkney Shetland has no pitchstone-bearing as- – it may have been perceived as imbued semblages. with symbolic meaning, representing contacts with ‘strange’ far-away places, The exchange of Yorkshire flint was dis- and probably obtained at a premium. On cussed by the author (Ballin 2011c) in Orkney, sites near the ritual centres may connection with his work on the lithic finds include approximately 50% Yorkshire 75

flint and some Arran pitchstone, where- implements probably having taken as sites near the island group’s periph- place ‘off the mountain’, that is, ei- ery – such as, Pool on Sanday and Links ther in the coastal parts of North of Noltland on Westray (Hunter 2007; Roe, or south of Rona’s Voe; it is un- Moore 2011) – contain no or very little certain whether this took place at Yorkshire flint and no pitchstone. specialized sites or at the domestic settlements. Conclusion • Although the exchange network of As explained above, the distribution of felsite may only cover Shetland it- felsite is complex and highly structured, self, prehistoric considerations regard with the dis-tribution of felsite indicating the distribution of felsite must ing multi-layered organization of the re- have included ‘the outside world’, duction process and exchange. It has with decisions having been made, been possible to identify several levels effectively restricting the exchange of organization: to the parent territory. • Within the North Roe quarry complex, two main forms of reduction The comparison of the felsite exchange have been recognized so far (more network with other early prehistoric net- may be defined in the future) – the works from Scotland and NW Europe quarrying of crude blocks/flakes, and identified important differences and the production of axehead and knife similarities. Three main forms of distri- preforms. butional expressions were identified, • It appears that the Neolithic quarri- namely: 1) simple linear regression; 2) ers defined the different forms of abrupt, or stepped, termination of the felsite as more or less suitable for the distribution; and 3) highly complex pat- production of axeheads and Shetland terns, usually characterized by multiple knives, respectively, and the concen- peaks and different forms of distribution trations of felsite waste may be sub- in different parts of the network. divided into 1) ‘tailing piles’, with waste having been deposited on top Simple linear regression seems to char- of, or alongside, the dykes, and 2) spe- acterize raw materials which were gen- cialized workshops for the produc- erally per-ceived in prehistory as function of axehead and knife preforms. tional, and the territories tend to be • The reduction of felsite included not relatively small, forming parts of larger only North Roe, but Shetland as a social territories. Abrupt, or stepped, whole, with the polishing of felsite termination of the distribution usually

Nordlige Verdener Shetlandsprojekt 76

characterizes raw materials which were where high-prestige central sites have perceived as emblematic, that is, they much ‘imported’ Yorkshire flint and defined particular social groups and Arran pitchstone, and ordinary do- their territories (‘our tribe uses felsite, mestic sites have none . rhyolite, etc., but other tribes don’t, as we have restricted the distribution of Combined, the above suggests that, de- this raw material’). And the more com- spite the limited geographical size of the plex exchange networks included sever- felsite exchange network, the reduction al social territories and several forms of and distribution of North Roe felsite was distribution, with people in the various highly and tightly organized, and the Ne- territories (at varying distances to the olithic society responsible for the dis- source) possibly perceiving the raw ma- semination of felsite probably quite so- terial in question differently: phisticated. However, compared to the 1) at the centre of the pitchstone net- neighbouring Orcadian social territory at work – that is, on Arran – pitchstone the same time – that is, during the Late was used al-most exclusively, and it Neolithic period – with its complex ritu- was probably perceived in emblem- al centres and a seemingly clear hierar- atic terms, defining people on Arran chical structure, the Shetland society as ‘those who use pitchstone’; off appears somewhat simpler. Where the Arran, objects in this raw material Orcadian Late Neolithic society almost are likely to have been seen as pre- has the appearance of a budding chief- cious exotica, with the value growing dom, the contemporary Shetland society towards the network’s periphery; has the appearance of a more egalitari- 2) the somewhat later Yorkshire flint net- an tribal society. However, much more work has a considerably larger core research needs to be carried out before area characterized by the almost ex- any firm conclusions can be made re- clusive use of this raw material; this garding any differences between these core area has a radius of c. 500 km, two interesting island groups. where the exclusive pitchstone centre had a radius of c. 50 km. It is possible that Yorkshire flint may have Bibliography been perceived in functional terms Ballin, T.B. 2005: Re-Examination of the in this area, whereas objects in this Quartz Artefacts from Scord of Brouster. raw material were perceived as pre- A lithic assemblage from Shetland and cious exotica at greater distances to its Neolithic context. Scottish Archaeo- the source, as indicated on Orkney logical Internet Reports (SAIR) 17. 77

[http://www.sair.org.uk/sair17/index. Ballin, T.B. 2011c: The Late Neolithic html]. lithic finds from Airhouse and Overhow- den, near the Overhowden Henge, Scot- Ballin, T.B. 2007: The Territorial Struc- tish Borders. Characterization and dis- ture in the Stone Age of Southern Nor- cussion of two unusual assemblages way, Wadding-ton, C. & K. Pedersen, K. and their parent sites. British Archaeo- (eds.), The Late Palaeolithic and Meso- logical Reports British Series 539. Ox- lithic of the North Sea Basin and Litto- ford: Archaeopress. ral, Proceedings from a Conference at the University of Newcastle-upon-Tyne, Ballin, T.B. 2011d: The post-glacial colo- 17 May 2003, Oxford: Oxbow Books: nization of Shetland – integration or iso- 114-136. lation? Evi-dence from lithic and stone assemblages, Mahler D. & Andersen C. Ballin, T.B. 2008: The lithic assemblage (eds.), Farming on the Edge: Cultural from Old Scatness Broch, Sumburgh, Landscapes of the North. Short papers Shetland. Unpublished report. from the network meeting in Lerwick, Shetland, September 7th-10th 2010. Co- Ballin, T.B. 2009: Archaeological Pitch- penhagen: The National Museum of stone in Northern Britain. Characteriza- Denmark. tion and interpretation of an important prehistoric source. British Archaeologi- Ballin, T.B. in prep.: The Laggan-Tormore cal Reports British Series 476. Oxford: lithic assemblage, Sullom Voe, Shetland. Archaeopress. Unpublished report.

Ballin, T.B. 2011a: The Felsite Quarries Ballin, T.B. & Johnson, M. 2005: A Me- of North Roe, Shetland – an overview, solithic Chert Assemblage from Glen- Davis V. & Edmonds M. (eds.), Stone Axe taggart, South Lanarkshire, Scotland: Studies III, Oxford: Oxbow Books: 121-130. Chert Technology and Procurement Strategies. Lithics 26: 57-86. Ballin, T.B. 2011: The felsite quarry complex of Northmaven: observations from Ballin, T.B., Barrowman, C., & Faithfull, a fact-finding mission to Shetland. In A. J. 2008: The unusual pitchstone-bearing Saville (ed.), Flint and Stone in the Neo- assem-blage from Blackpark Plantation lithic Period. Neolithic Studies Group East, Bute. Transactions of the Buteshire Seminar Papers 11, 62-81. Oxford: Ox- Natural History Society XXVII: 23-51. bow Books.

Nordlige Verdener Shetlandsprojekt 78

Fojut, N. forthcoming: The Stourbrough Renfrew, C. 1977: Alternative Models for Hill Hoard and other Polished Stone Exchange and Spatial Distribution, T.K. ‘Knives’ from Shetland. Earle, T. K., & EricsonJ. E. (eds.), Exchange Systems in Prehistory. Studies in Archae- Hedges, J.W., & Parry, G.W. 1980: ‘A ology, 71-90. New York: Academic Press. Neolithic multiple burial at Sumburgh Airport, Shetland’, Glasgow Archaeo- Ritchie, R. 1992: Stone Axeheads and logical Journal 7: 15-26. Cushion Maceheads from Orkney and Shetland: Some Similarities and Con- Hunter, J. 2007: Excavations at Pool, trasts, Sharples, N. & A. Sheridan, A. Sanday. A multi-period settlement from (eds.), Vessels for the Ancestors. Essays Neolithic to Late Norse times. Investiga- on the Neolithic of Britain and Ireland in tions in Sanday, Orkney 1. Kirkwall / Edin honour of Audrey Henshall, 213-220. Ed- burgh: Orcadian Ltd. / Historic Scotland. inburgh University Press, Edinburgh.

Högberg, A., & Olausson, D. 2007: Scan- Schneiderman-Fox, F., & Pappalardo, dinavian Flint – an Archaeological Per- A.M. 1996: A Paperless Approach To- spective. Aarhus: Aarhus Universitets- ward Field Data Collection: An Example forlag. from the Bronx. Society for American Archaeology Bulletin 14.1. [http://www. Kindgren, H. 1991: Kambrisk flinta och saa.org/publications/saabulletin] etniska grupper i Västergötlands sen- mesolitikum, Browall, H., Persson, P and Scott, D.D., & Thiessen, T.D. 2005: Catli- Sjögren, K. G. (eds.), Västsvenska sten nite Extraction at Pipestone National åldersstudier, vol. 1. Gotarc, serie c, Monument, Minnesota: Social and Tech- Arkeologiska skrifter 8, 71-110. Göteborg: nological Implications. Topping, P., & Ly- Institutionen för arkeologi, Göteborgs nott, M. (eds.), 2005: The Cultural Land- universitet. scape of Prehistoric Mines, Oxford, Oxbow Books: 140-154. Moore, H. and Wilson, G. 2011: Shifting Sands. Links of Noltland, Westray: In- Topping, P. 2005: Shaft 27 Revisited: An terim Report on Neolithic and Bronze Ethnography, Topping, P., & Lynott, M. Age Excavations, 2007-09. Archaeology (eds.), 2005: The Cultural Landscape of Report 4. Edinburgh: Historic Scotland. Prehistoric Mines, Oxford, Oxbow Books: 63-93. 79

Axes from islands: the role of stone axeheads from insular sources in the Neolithic of Ireland and Britain

Gabriel Cooney

Introduction discuss the significance of island stone A notable feature of the Neolithic (4000- quarries and the distribution of objects 2500 cal BC) of northwest Europe is the from these sources as an aspect of the exploitation of lithic sources on islands introduction of farming and the develop- for the production of stone axeheads ment of Neolithic societies in the Irish and other artifacts. Examples include Sea Zone and North Atlantic. Relevant Hespriholmen in western Norway (Als issues to be considered include the date aker 1987; Bergsvik 2006), Helgoland at which organized quarrying commenced, (Germany) off the coast of Denmark (e.g. production processes and the spatial lo- Wentink 2006) and sites on the Channel cation of different stages of production. Islands (Patton 1991). This paper focuses These issues will be linked to the exami on work being carried out by the author nation of the extent and character of the and colleagues on three such islands; distribution of axeheads from the sourc- Lambay (source: porphyry) in the Irish es and the social networks involved. Sea off the east coast of Ireland, Rathlin (source: porcellanite) in the North Chan- Lambay, local lives nel off the northeast coast of Ireland and wider landmarks and the island group of Shetland (source: Lambay is located about 11 km north of felsite) between the North Atlantic and the peninsula of Howth, which forms the the North Sea. The evidence from quarry northern edge of Dublin Bay, and 8 km sites on these islands will be used to from the nearest part of the Dublin coast-

Nordlige Verdener Shetlandsprojekt 80

Fig. 1 (left): Lambay porphyry (porphyritic andesite) , a piece of outcrop and an axe produced from the same material. Photo: Gabriel Cooney

Fig. 2 (right): Excavated quarry face and surface of working floor with debitage, Eagle’s Nest site, Lambay. Photo: Gabriel Cooney

line. It is the largest island off the east We can say something about the lives coast of Ireland and has a dramatic sky- that Neolithic people lived on Lambay line, created by the fact that the island from surface collections of lithics (Dolan is largely volcanic in geological origin and Cooney 2010). The storm beaches (Stillman 1994), added to by a large cairn on the west and south coast appeared on the highest point of the island, Knock- to provide the most abundant source of bane (the white cairn). Excavation has flint, which was the dominant lithology been carried out at a Neolithic axe quarry used for small stone tools. The heaviest site known as the Eagle’s Nest, close to concentrations of lithics occurred in the the centre of the island (e.g. Cooney 1998; same general area as the flint sources, 2005; 2009). There is clear evidence on although there was a widespread distri- the site for material culture and events bution of material across the island. In- in the Neolithic that seem both to evoke deed the largest assemblage came from the particular character of Lambay as an a terrace on the east-facing slope of the island but that also link it to a wider upland core of the island. It would also world. Examining such evidence one is appear that there were persistent places reminded of Broodbank’s (2000: 363) de- that were used in the island landscape, scription of island people as people who as on the southern coast near the head- are neither entirely different from the land known as Black Point. The diagnos- rest of the world nor yet wholly similar. tic artifacts were dominated by Neolithic 81

material but there were a couple of pos- Fig. 3: sible Early Mesolithic cores and a Late Middle Neo- lithic setting Mesolithic butt-trimmed flake. of stone focused on a What this material helps us to appreci- large slab, Eagle’s Nest ate is the human context of the use of site, Lambay. the Eagle’s Nest quarry site during the Photo: Gabriel Neolithic. Here a distinctive medium- Cooney grained volcanic rock, porphyry, or porphyritic andesite, was exploited for the production of stone axeheads (Fig. 1). Some of the contemporary features are The primary process of production was reminiscent of the settings outside pas- by hammering and pecking followed by sage tombs. One of the notable features grinding. All stages of axe production of the material culture is the deposition took place at the site. Radiocarbon dates of jasper; broken and tested pebbles indicate that quarrying started around and a small number of beads and pen- 3800 cal BC (Early Neolithic) and there dants, the latter of a type that are typi- is evidence of the (episodic?) use of the cally found in deposits in passage tombs site well into the later part of the fourth (e.g. Eogan 1986). The evidence indicates millennium cal BC (Middle Neolithic) that jasper outcrop veins and beach peb- (Cooney et al. 2011). Quarrying activity bles were being worked, either at the (Fig. 2) is complemented by deliberate site or elsewhere on the island. Deposits deposition, both in the quarry areas and of beach gravel at the site are another on the floor of a small valley between indicator of the deliberate deposition of two worked outcrops. On this valley material brought up from the coast. Ulti- floor there is a sequence of features that mately this buildup of material resulted begins with the digging and filling of in the creation of a low cairn, at maxi- pits, some which appear to be deliber- mum about 10m in diameter. ately cut into by later pits. Then there is a switch to the placement on the ground of It is attractive to think of the Eagle’s features (Fig. 3) and a range of cultural Nest site as a place where porphyry was material was deposited in a number of worked and deposited and where other events, for example at some point in the material was brought and also deposited. Middle Neolithic a hoard of objects in- In these actions the connections between cluding a pestle macehead, a porphyry a range of material culture, and the peo- roughout and axehead was deposited. ple involved, was reworked. The place-

Nordlige Verdener Shetlandsprojekt 82

ment of material on or under the ground Neolithic activities on Lambay cannot may have been viewed as making link- be seen in isolation. What we are see- ages through a powerful conduit to the ing are the links between activities on a realm of the supernatural (La Motta and specific island and the wider cultural Schiffer 2001). In the deposition of por- setting in which they took place. In the phyry axes and roughouts we see the Middle and Late Neolithic a term which beginning and the end of the biographical Cunliffe (2001) has used for this wider cycle of objects happening at the same cultural setting is the Boyne-Orcadian location. Utilizing Bradley’s (2000) test of axis (see also discussion in Cooney the significance of natural places, his three 2000: 224-8). As a specific example the major signifiers occur at the Eagle’s Nest pestle type macehead (and there are site: distinctive deposits, the embellish- fragments of others) from the Eagle’s ment of the landscape and artifacts made Nest site can be compared with those at special locations. It should also be found in settlement and tomb contexts noted that the majority of porphyry axe- in Orkney and with the only example heads that can be said on the basis of found in secure archaeological context petrological and geochemical analysis in Ireland, from outside the outer sill- to definitely originate from the Eagle’s stone in the passage of the western Nest quarry were actually found in the tomb structure under the main mound at excavation at the quarry. Knowth (Simpson and Ransom 1992, 227). Interestingly in the surface collect- Moving away from the quarry site it is ed material from close to the harbour on worth noting that there are quarried por- the western side of the island there was phyry pieces in the makeup of the cairn a worked piece of pitchstone from the at Knockbane, the most notable prehis- the island of Arran (see Ballin 2009), toric landmark on the island. Knockbane providing another sign of a link with the also takes us out to the wider world, it is Irish Sea world. placed to be seen. To the northeast on a clear winter’s day the Isle of Man is vis- What we have at Lambay then in the ible. Looking north from Knockbane there form of the quarrying and depositional is a distant view to Slieve Gullion (with activity and the construction of Knock- a passage tomb on the summit) and the bane is the material manifestation of Carlingford and Mourne mountains (in- what Robb (2001: 196) has called the cluding Slieve Donard with the highest ‘reworking of a regional symbolic herit- hill-top passage tomb in Ireland). Given age into a local, cosmologically ground- these kinds of links it is obvious that the ed identity’. 83

A centre of island life, axe production on Rathlin Island

Gabriel Cooney, Stephen Mandal, Emmett O’Keeffe and Graeme Warren

Rathlin lies just over 10 km off Fair Head at the northeast corner of Ireland. An outstanding feature of the range of evidence for Neolithic activity on the island is the quarrying and production of porcellanite at Brockley (Fig. 4). Porcellanite accounts for at least 50% of the total number of Irish stone axeheads (Cooney and Man- dal 1998), with a significant number known from Britain (Sheridan 1986; Sheridan et al. 1992: 410; porcellanite is labelled as Group IX in the British scheme for the petrological classification of stone Fig. 4: View of the porcellanite outcrop at Brockley from the south. axeheads and their sources). It takes a Photo: Emmett O’Keeffe. conchoidal fracture when struck, hence the chaîne opératoire by which porcel- et al. 1997). The results of this work have lanite was shaped and transformed from demonstrated that the traditional focus raw material to cultural product can be on Tievebulliagh, which has also been the followed (Bradley and Edmonds 1993; subject of greater archaeological research Mallory 1990). (e.g. Jope 1952; Morey and Sabine 1952; Sheridan 1986; Mallory 1990), needs to There are two known sources for porcel- be re-assessed. This is supported by the lanite axe production, at Tievebulliagh extent of quarrying at Brockley, and of near Cushendall in Co. Antrim and Brock- the evidence for the processing and use ley on Rathlin. Cooney and Mandal of porcellanite from across the island. (2000: 52–55) have reviewed the petro graphy of porcellanite axes and the The porcellanite at Brockley was formed sources. Trace-element geochemical (XRF) by the metamorphism of an interbasaltic analysis has been applied to differenti- horizon laterite in situ by a large dolerite ate the two known sources and products intrusion (Dawson 1951). At the foot of from them (Meighan et al. 1993; Mandal the south-east side of the rock face, two

Nordlige Verdener Shetlandsprojekt 84

Fig. 5: Plan and section of galleries 1 and 2, Brockley. Drawing: Conor Brady

Galleries 1 and 2 resemble the primary extraction sites at Top Buttress, Pike of Stickle, Great Langdale in Cumbria. These were excavated (Bradley and Ed- monds 1993) as part of the project to understand the process of Neolithic extraction of volcanic tuff for the production of Group VI axes, the single most important stone source for axeheads in Brit- ain. While there has been no excavation at Brockley itself, it is clear from the evidence of porcellanite objects that the primary purpose of quarrying was the production of roughouts for stone axeheads. Flaking was the predominant primary treatment, with minor evidence of pecking and cleaving (Cooney and Man- dal 1998: 60).

The most widely occurring evidence for the process of axe production comes from the very large number of porcellan- extraction sites (galleries 1 and 2; Figs. ite roughouts in museums and other col- 4 and 5), some 5 m apart, are visible lections. Analysis of the Irish Stone Axe where rock extraction has created clear Project (ISAP) database suggests that overhangs. They face almost due south there are a number of recurring rough- and are clearly the result of quarrying out types. A feature that should be not- the lowest part of the porcellanite out- ed is the relatively high percentage of crop. Conchoidal fracture scars can be porcellanite axeheads with oblique- seen on the roofs and sides of both. Fif- shaped butts. While some 9% of Irish teen metres to the north-east of gallery 1 stone axeheads in general have this is what appears to be an almost com- form of butt, the figure rises to 16% for pletely infilled quarry area (gallery 3). those of porcellanite (Cooney and Man- 85

dal 1998: 61). It is clear that that this vations at Knockans (Conway 1996) and feature had stylistic value, as many of Craigmacagan (Logue 2005) have re- the finest polished porcellanite axe- vealed significant evidence for debitage, heads have oblique butts (Fig. 6). roughouts and finished axeheads. The evidence for the working of porcellanite There are significant numbers of porcel- from these and older excava- lanite roughouts (over 250) and finished tions matches the range of axeheads (over 100) known from Rathlin and these along with porcellanite debit- Fig. 6: Porcellanite roughout age have been found at many different from Rathlin with an oblique butt. places (Fig. 7). In addition, recent exca- Photo: Wes Forsythe.

Fig. 7: Distribution of porcellanite roughouts and axeheads on Rathlin. Drawing: Emmett O’Keeffe

Nordlige Verdener Shetlandsprojekt 86

forms in museum collections. Hence the the 32 axeheads analysed (Mandal et al. extraction site at Brockley was linked 1997: table 2), the five from Rathlin all into a process of axe production that was have Sr values correlating with the Brock spread widely across the island. The ley source. Of the 27 other axeheads, 13 distribution patterns suggest most of the fall within the Brockley source field and activity took place to the east of Brock- a further two probably originated from ley. The evidence from excavations indi- the Brockley source. This strongly sup- cates that following on from quarrying, ports the view that the importance of all stages of axe production took place the Brockley source has been greatly at different locations on the island. underestimated in the literature. About 30% of the porcellanite axeheads from Rathlin have been ground and pol- Recognising that it is only a very small ished. Given the number of ground and sample, looking at those axeheads that polished axeheads now known from the appear to have originated from Brockley island, it is clear that there was no spatial and have provenances off the island al- separation of the primary (flaking) and lows us to make further comments about secondary (grinding/polishing) stages of the character of porcellanite stone axe production. This indicates that a much production on and distribution from the higher percentage of finished axeheads island of Rathlin. They provide support was exported from the island than was for the suggestion by Sheridan (1986: previously believed, based on a model 22) that some of the material from Rath- emphasising the transport of roughouts lin was transported by sea along the to coastal locations on the mainland. north coast of Ireland. The identified Brockley artefact that is the greatest That axeheads from the Brockley source distance from the source is an axehead were transported off the island has been from Clonsilla, Co. Dublin (NMI confirmed by trace-element geochemi- 1942:715). This is 18.7 cm in length. and cal (XRF) analysis of samples taken from argues against a simple down-the-line the two sources at Tievebulliagh and exchange process which would tend to Brockley and from a selection of porcel- see small axeheads dominant with in- lanite axeheads (Meighan et al. 1993; creasing distance from the source as the Mandal et al. 1997). Tievebulliagh and axeheads are re-worked (see discussion Brockley samples have different strontium in Chappell 1987: chapter XI). Converse- levels, hence it is now possible, using Sr ly there are porcellanite roughouts from as a discriminant, to relate porcellanite Rathlin that are only 9-10 cm in length, axeheads to their respective sources. Of indicating that small axeheads as well 87

as larger ones were being produced on objects are roughouts. What is interest- the island. ing here are the physical connections made between the different stages of axe On Rathlin there are a small number of production and an axehead made of a axeheads made from non-local stone different lithology. Bradley (1990: 33) has sources; two gabbro axeheads and two commented that archaeologists actually shale axeheads which could have been observe only two stages in the life cycle produced from a number of sources in of an artifact: its production and final de Northern Ireland, but not on the island. position, we have to work much harder One of the shale axeheads from Rathlin to actually understand how axes were appears to have come from a hoard of five used and moved. What we can observe axeheads (Briggs 1988: 7), the other four from the archaeological record of axe being of porcellanite. The shale axehead production on Rathlin is that it was both is complete and is ground and polished, an important aspect of island life and that as are two of the porcellanite axeheads, it provides us with a physical, material one complete, one the lower portion of reminder of a wider world of contacts an axehead. The other two porcellanite and movement to and from the island.

Objects for an island world, North Roe Felsite, Shetland

Gabriel Cooney, Torben Ballin and Will Megarry

While not widely known or discussed in were exploited. Felsite dykes, as mapped the archaeological literature, the felsite by Phemister et al. (1952) are orientated quarries of the North Roe peninsula, north-south and visually (as grey, blue, mainland Shetland probably represent bluish-green or purple) stand out against the most extensive and best-preserved the red granite bedrock (Fig. 8). Ritchie Neolithic quarry complex in Britain or (1968; 1992) noted the significance of Ireland. As detailed by Ballin (2011a) the the exploitation of these dykes for the term felsite is used to refer to a number production of axeheads and Shetland of related rock types. Ritchie (1968) sin- knives. Varieties of felsite are visually gled out riebeckite felsite, due to its spectacular when polished. Ballin’s spectacular appearance but it would ap- (2011a) examination of felsite imple- pear that a number of forms of felsite ments in the Shetland Museum suggests

Nordlige Verdener Shetlandsprojekt 88

Fig. 8: The contrast between granite and felsite outcrops, North Mavine, Shetland. Photo: Gabriel Cooney

that axeheads were made from gener- Ballin (2011a; 2011b) has undertaken ally homogeneous forms of felsite, with- several short reconnaissance survey stints. out spherulites, but with small phenoc- This work indicates that different types of rysts of quartz and feldspar. The knives felsite were used in the production of are more striking in appearance, with tools. There are areas where felsite appear above-average size spherulites, but to have been exploited extensively and mostly containing few phenocrysts. others where activity seems to have been Hence the aesthetic appearance of the more sporadic. Different kinds of quarry artifacts appears to have been a very activity areas have been recognised: quar- significant factor across the Shetland ries, tailing piles with quarried material, archipelago. block reduction sites and work sites where 89

tween the two main quarry complexes). An important factor to be considered is the presence of blanket peat over much of the complex, particularly the lower- lying areas. There are locations where the peat has been eroded and felsite can be seen but it is difficult to characterize the nature of this material. More broadly the issue is the amount and character of evidence that may be con- cealed by the blanket peat cover.

Preliminary analysis of museum collections indicates that the distribution of felsite products was widespread within but restricted to the Shetland archipelago (see Ballin 2011c: 38-9). Combined with the exceptional quality of the evidence for quarrying and production in North Roe this geographic context facilitates following the cultural life of felsite objects through extraction, production, use- life and deposition, in some cases as spectacular caches or hoards and in other cases in houses, where they mainly occur the tool pre-forms or rough-outs were pro- in fragmentary or re-used form. The dis- duced. Potentially related features such tribution and context of these artifacts as small shelters and cairns can be recog- can be taken as a proxy for the amount nized and at the Beorgs of Uyea there is and nature of exchange and contact with- the well-known lintelled gallery, with one in and between island communities in lateral side formed by a worked felsite Shetland and how this changed over time. dyke (Scott and Calder 1952). Three dis- This evidence provides a critically im- tinct zones can be recognized in the quar- portant research opportunity to examine ry complex; The Beorgs (of Uyea) / Pet- the dynamics of island life and compare tadale Water, Midfield (east of Ronas Hill) and contrast it with Neolithic societies and the central Lakelands (the area be- elsewhere in Europe.

Nordlige Verdener Shetlandsprojekt 90

Fig. 9: Palimpset with multiple episodes of quarried blocks, Beorgs of Uyea. Photo: Gabriel Cooney 91

During September 2011, following up on orgs of Uyea, for example, is defined by earlier reconnaissance work, selected extensive exploitation, probably over a areas of North Roe were inspected as prolonged period of time, and the de- part of the definition and planning of a posits of quarrying waste have a clear project investigating Neolithic felsite palimpsest appearance (Fig. 9). By con- quarrying in North Roe. The foci of this trast, the Midfield summit, or ridge, is project will be: felsite quarrying tech- characterized by the presence of many nology and organization; felsite blank, discrete, probably single-event, work- preform and tool technology and organi- shops and clusters of workshops. zation; felsite exchange within North Roe/Shetland, and beyond; dating the Looking at Midfield in more detail, it exploitation of felsite; and finally, build- was possible to define two main areas ing on the evidence-base of the project, of activity, associated with two parallel the social role of felsite in Neolithic felsite dykes. Midfield 1 (western dyke) Shetland. is characterized by two main quarry pits and one main large workshop, probably As previous years’ archaeological activ- representing repeated exploitation (al- ity in North Roe (Ballin 2011a; 2011b) though not to a degree comparable to had shown that felsite dykes in the cen- that seen at the Beorgs), relating to tral (Lakelands) parts of the peninsula combined axehead and knife production. had been prospected by prehistoric peo- Midfield 2 (eastern dyke), on the other ple, but not subjected to organized ex- hand, is characterized by four or five ploitation, work focused on selected lo- quarry pits along a felsite outcrop (Fig. cations on the Beorgs of Uyea ridge 10), with discrete, probably single-event (northern North Roe) and along the elon- workshops, located on either side of the gated summit of Midfield (southern dyke. This eastern dyke follows the land- North Roe) east of Ronas Hill. Scrutiny scape contours, and workshops defined of axehead and knife rough-outs and by coarse waste from axehead produc- production waste, as well as the differ- tion were found on the eastern (down- ent types of felsite available in the se- slope) side of the dyke, whereas work- lected areas, indicates that axeheads shops defined by finer waste from knife and Shetland knives were manufactured production were found on the western in both locations. However, it was also (up-slope) side of the dyke (Fig. 11). possible to characterize the two areas as having different potential value for The Midfield 2 quarry pits and work- archaeological investigation. The Be- shops have been identified as an initial

Nordlige Verdener Shetlandsprojekt 92

Fig. 10: Quarry pits along the eastern dyke at Midfield, looking south. Photo: Gabriel Cooney 93

Fig. 11: Knive workshop at eastern dyke, Midfield. Photo: Gabriel Cooney

focus of the project. Understanding dis- As mentioned above the aim of the project crete activity areas will facilitate analy- is to understand the role of this distinc- sis and interpretation of other areas of tive local island stone source – felsite, the quarry complex which have a pal- which Neolithic people physically and impsest character and the wider role of culturally transformed into axeheads and felsite artifacts in Neolithic Shetland. the highly distinctive Shetland knives. It is Key objectives of the project are the de- recognized that Shetland is the northern tailed survey, mapping and characteri- most part of Europe where farming was sation of the quarry complex. This will practiced during the Neolithic (3800- be combined with an analysis of muse- 2500 cal BC). In the archipelago there is um collections. A project GIS (Geograph- widespread evidence of the stone houses, ical Information Systems) is being de- settlements, field systems and tombs that veloped and is seen as essential for the early farmers built and used. The quar- effective and accurate survey of the rying, production and use of felsite ob- quarry complex. The multi-scalar and re- jects from North Roe provides an oppor- flexive character of the GIS means that tunity to explore the links between these it can facilitate research at differing different nodes of Neolithic activity and to scales and assist in research design. understand how felsite quarrying form ed part of Neolithic society in Shetland.

Nordlige Verdener Shetlandsprojekt 94

Stone: Bringing the notion of special places and inter-connection together

Gabriel Cooney

It is now recognised that for prehistoric ture between notions of islands as fixed, societies stone was ‘symbolically mean- bounded places and the fluidity and in- ingful, ritually powerful and deeply inter-connectivity of islands and main- terwoven into not just economic and lands created by the movement of peo- material, but also social, cosmological, ple, the central importance and role of mythical, spiritual and philosophical as- the sea and contact with other places in pects of life’ (Boivin 2004: 2). Recogni- island life (see discussion in Van de tion that objects can be animate and re- Noort 2011). Study of the use of stone garded as active makes it useful to think by island communities provides us with of them from a biographical perspective the opportunity to explore both these el- (Kopytoff 1986; Gosden and Marshall ements of island life. Objects of stone 1999; Davis and Edmonds 2011). The brought from or to islands literally carry character of the archaeological record their sense of place and history with at and from sites where stone was quar- them. The movement of material be- ried and worked has potential to con- tween locations and sources provides tribute to wider issues, for example how us with the opportunity to think about materials were engaged with in specific exchange systems. social and historical contexts (Ingold 2007; O’Connor and Cooney 2009: xxii; Ballin (this volume) has suggested that Cooney 2011) we should think of the exchange of felsite as having an emblematic social While understanding of the archaeolog- function within Shetland and that more ical record is based on quite different broadly the main factor influencing the histories of research in each case, the exchange of raw materials during the three island case studies discussed Neolithic was whether they had func- above demonstate both the significance tional and/or symbolic values. Looking of the use of stone and its place in the at Lambay porphyry the very restricted Neolithic world and how we can use is- distribution of objects from the Eagle’s land lithic sources to address the com- Nest site allied to the visually distinc- plementaries recognised in the litera- tive character of the axeheads seems to 95

support a symbolic interpretation. While in Shetland or whether, like on Lambay the density of porcellanite from Rathlin and Rathlin the quarrying and procure- and Tievebulliagh does generally de- ment of axes from special sources may crease with distance from the known have defined from the start of what it quarries, there are some concentrations was to be ‘Neolithic’ (Cooney 2008). at distance from the sources, suggesting that it was an important component Acknowledgements in contact networks both within Ireland The section on Rathlin owes a lot to the and across the Irish Sea. work of my colleagues Stephen Mandal, Emmett O’Keeffe Graeme Warren and Going back the critical issue of the rela- the section on North Roe is heavily reli- tionship between these exchange net- ant on the work of Torben Ballin and works and the establishment of the Ne- also that of Will Megarry, our collabora- olithic, it will be clear that our state of tor on the North Roe Felsite Project. My knowledge is quite different for the thanks to all my colleagues who contrib- three islands. In the case of Lambay uted to the paper and to the editor for there is direct evidence of quarrying his patience! from early in the Neolithic. There are no dates directly related to porcellanite production, but porcellanite axeheads Bibliography and related debitage are known from key early Neolithic sites in Ireland such Alsakar, S. 1987: Bomlo – Steinalderens as Magheraboy (Danaher et al. 2007) rastoffsentrum pa Sorvestlandet. Ber- and Donegore Hill (Mallory et al. 2011) gen Historisk Museum, Universitetet i causewayed enclosures suggesting that Bergen. Bergen. porcellanite production was a feature of how relationships were developed and Ballin, T. 2009: Archaeological pitch- maintained within and between the ear- stone in Northern Britain, characteriza- liest farming communities in Ireland. In tion and interpretation of an important both cases it is clear that there was sig- prehistoric source, British Archaeologi- nificant, continued use of the sources in cal Reports British Series 476. Oxford. the Middle Neolithic. In this context one of the key questions to be asked about Ballin, T. 2011a: The felsite quarries of the North Roe felsite quarry complex is North Roe, Shetland – An overview. whether, as the current view suggests it Stone Axe Studies III. Davis, V. and Ed- forms part of the established Neolithic monds, M. (eds.), Oxbow, Oxford: 121-30.

Nordlige Verdener Shetlandsprojekt 96

Ballin, T. 2011b: The Felsite Quarry Com- Briggs, C.S. 1988: Stone resources and plex of Northmaven. Observations from implements in prehistoric Ireland: a review. a fact-finding mission to Shetland. Flint Ulster Journal of Archaeology 51: 5–20. and Stone in the Neolithic Period. Saville, A. (ed.), Oxbow, Oxford: 62-81. Broodbank, C. 2000: An Island Archaeol- ogy of the Early Cyclades. Cambridge Ballin, T. 2011c: The Post-Glacial Coloni- University Press, Cambridge. zation of Shetland – Integration or Isola- tion. Farming on the edge: Cultural land- Chappell, S. 1987: Stone Axe Morphol- scapes of the North. Mahler, D. and ogy and Distribution in Neolithic Britain. Andersen, C. (eds.), National Museum British Archaeological Report British Se- of Denmark, Copenhagen: 32-43. ries 177, Oxford.

Bergsvik, K.A. 2006: Ethnic Boundaries Conway, M. 1996: Excavations at ‘Shan- in Neolithic Norway. British Archaeolog- dragh’, Knockans, Rathlin Island, 1994- ical Reports, International Series 1554, 1996. A preliminary statement (unpub- Oxford. lished summary account of excavations).

Boivin, N. 2004: From veneration to ex- Cooney, G. 1998: Breaking stones, mak- ploitation: human engagement with the ing places: the social landscape of axe mineral world. Soils, Stones and Sym- production sites. Prehistoric Ritual and bols: Cultural Perceptions of the Mineral Religion. Gibson, A. and Simpson, D.D.A. World. Boivin, N. and Owoc, M.A. (eds.), (eds.). Sutton, Stroud: 108-18. UCL Press, London: 1-29. Cooney, G. 2000: Landscapes of Neolith- Bradley, R.J. 1990: The Passage of Arms, ic Ireland. Routledge, London. Cambridge University Press, Cam- bridge. Cooney, G. 2005: Stereo Porphyry: Quar- rying and Deposition on Lambay Island, Bradley, R.J. 2000: An Archaeology of Ireland. The Cultural Landscape of Pre- Natural Places. Routledge, London. historic Mines. Topping, P. and M. Lynott (eds.). Oxbow Press, Oxford: 14–29. Bradley, R.J. and Edmonds, M. 1993: In- terpreting the axe trade: production and Cooney, G. 2008: Engaging with Stone: exchange in Neolithic Britain, Cambridge Making the Neolithic in Ireland and University Press, Cambridge. Western Britain. Between foraging and 97

farming: an extended broad spectrum of Cunliffe, B. 2001: Facing the Ocean. Ox- papers presented to Leendert Louwe ford University Press, Oxford. Kooijmans. Fokkens, H., Coles, B. J., Van Gijn, A. l., Kleijne, Jos P. H., Ponjee, Danaher, E. (ed.) 2007: Monumental Be- H. and Slappendel, C. G. (eds.). Analecta ginnings: The archaeology of the N4 Sli- Praehistorica Leidensia 40: 203-214. go Inner Relief Road. NRA Scheme Mon- ograph 1, Dublin. Cooney, G. 2009: The prehistory of Lam- bay, a long view. Axes, warriors and Dawson, J. 1951: The Brockley dolerite windmills: Recent archaeological dis- plug and the Church Bay volcanic vent, coveries in north Fingal. Baker, C. (ed.). Rathlin Island, Co. Antrim. Irish Natural- Fingal County Council, Swords: 8–22. ist Journal 10: 156–161.

Cooney, G. (ed.) 2011: New approaches Davis, V. and Edmonds, M. (eds.) 2011. to stone mines and quarries: materials Stone Axe Studies III. Oxbow, Oxford. and materiality. World Archaeology 43.2. Dolan, B. and Cooney, G. 2010: Lambay lithics: the analysis of two surface col- Cooney, G. and Mandal, S. 1998: Irish lections from Lambay, Co. Dublin. Pro- Stone Axe Project Monograph I. Word- ceedings of the Royal Irish Academy well, Dublin. 110C: 1-33.

Cooney, G. and Mandal, S. 2000: The Eogan, G. 1986: Knowth and the Pas- Irish Stone Axe Project: Sources for- sage Tombs of Ireland. Thames and Hud- stone axes in Ireland. Krystalinikum 26: son, London. 45–55. Gosden, C. and Marshall, Y. 1999: The Cooney, G., Bayliss, A., Healy, F., Whit- cultural biography of objects. World Ar- tle, A., Danaher, E., Cagney, L., Mallory, chaeology 31: 169-78. J., Smyth, J., Kador, T. and O’Sullivan, M. 2011: Chapter 12: Ireland. Gathering Jope, E.M. 1952: Porcellanite axes from Time: Dating the Early Neolithic Enclo- factories in northeast Ireland: Tievebul- sures of Southern Britain and Ireland. liagh and Rathlin. Ulster Journal of Ar- Whittle, A., Healy, F. and Bayliss, A. chaeology 15: 31–55. (eds.). Oxbow, Oxford: 562-669.

Nordlige Verdener Shetlandsprojekt 98

Ingold, T. 2007: Materials against mate- Meighan, I.G., Jamison, D.D., Logue, riality. Archaeological Dialogues 14: 1-38. P.J., Simpson, D.D.A., Rogers, G. Man- dal, S. and Cooney, G. 1993: Trace ele- Kopytoff, I. 1986: The cultural biography ment and isotopic provenancing of North of things: commoditization as process. Antrim porcellanites: Portrush-Tievebul- The Social Life of Things: Commodities liagh-Brockley (Rathlin Island). Ulster in Cultural Perspective, Appadurai, A. Journal of Archaeology 56: 25–30. (ed.). Cambridge University Press, Cam- bridge: 4-91. Morey, J.E. and Sabine, P.A. 1952: A petrographical review of the porcellan- Lamotta, V.M. and Schiffer, M.B. 2001: ite axes of North east Ireland. Ulster Behavioural archaeology: towards a Journal of Archaeology 15: 56–60. new synthesis. Archaeological Theory Today, Hodder, I. (ed.). Polity Press, Cam- O’Connor, B. and Cooney, G. 2009: Intro- bridge: 14-64. duction: materialitas and the significance of stone. Materialitas: Working Logue, R. 2005: Excavations at Craig Stone, Carving Identity. macagan Townland, Rathlin Island, County Antrim. Data Structure Report. O’Connor, B, Cooney, G. and Chapman, Centre for Archaeological Fieldwork, J. (eds.). Prehistoric Society/Oxbow, Ox- The Queen’s University of Belfast. Belfast. ford: xxi-v.

Mallory, J.P. 1990: Trial excavations at Patton, M. 1993: Statements in Stone: Tievebulliagh, Co. Antrim. Ulster Jour- Monuments and Society in Neolithic nal of Archaeology 52: 15–28. Brittany. Routledge, London.

Mallory, J.P., Nelis, E. and Hartwell, B. Phemister, J., Harvey, C.O. and Sabine, 2011: Excavations on Donegore Hill, Co. P.A. 1952: The riebeckite-bearing dikes Antrim. Wordwell, Dublin. of Shetland. Mineralogical Magazine 29: 359-373. Mandal, S., Cooney, G., Meighan, J. and Jamison, D.D. 1997: Using geochemis- Ritchie, P.R., 1968: The stone implement try to interpret porcellanite axehead pro- trade in third-millennium Scotland. Stud- duction in Ireland. Journal of Archaeo- ies in Ancient Scotland. Coles, J.M. and logical Science 24: 757-63. Simpson, D.D.A. (eds.). Leicester Uni- versity Press, Leicester: 119-136. 99

Ritchie, P.R. 1992: Stone axeheads and Sheridan, A., Cooney, G. and Grogan, E. cushion maceheads from Orkney and 1992: Stone axe studies in Ireland. Pro- Shetland: some similarities and con- ceedings of the Prehistoric Society. Pre- trasts. Vessels for the Ancestors. Sharp- hist. 58: 389-416. les, N. and Sheridan, A. (eds.). Edinburgh University Press: 213-220. Simpson, D.D.A. and Ransom, R. 1992: Maceheads and the Orcadian Neolithic. Robb, J. 2001: Island identities: ritual, Vessels for the Ancestors. Sharples, N. travel and the creation of difference in and Sheridan, A. (eds.). Edinburgh Uni- Neolithic Malta. European Journal of versity Press: 221-243. Archaeology 5: 24-41. Stillman, C. 1994: Lambay, an ancient Scott, L.G. and Calder, C.S.T. 1952: Notes volcanic island in Ireland. Geology To- on a chambered cairn, and a working day 62: 62-67. gallery, on the Beorgs of Uyea, North- maven, Shetland. Proceedings of the Van de Noort, R. 2011: North Sea Ar- Society of Antiquaries of Scotland 86: chaeologies: A maritime biography 171-177. 10,000 BC – AD 1500. Oxford University Press, Oxford. Sheridan, A., 1986: Porcellanite artifacts: a new survey. Ulster Journal of Wentink, K. 2006: Ceci n’est pas une Archaeology 49: 19–32. hache: Neolithic Depositions in the Northern Netherlands. Karsten Wentink, Leiden.

Nordlige Verdener Shetlandsprojekt 100

Multi-period construction of megalithic tombs – and the megalithic tombs in Shetland.

Flemming Kaul

Introduction monumental architecture. The landscapes Around 4000 BC the agrarian economy were changed into ritual landscapes swiftly spread over large areas of North- – from natural landscapes to cultural west Europe. Both in North Germany, landscapes. Today we can still see the South Scandinavia and on the British megalithic tombs in our landscapes, stand- Isles and Ireland an economy based on ing as the most solid testimony of these hunting, fishing and gathering was re- changes that took place almost 6000 placed by farming and animal husbandry. years ago. This change had of course far reaching consequences for the Stone Age socie- Almost everywhere in Western Europe ties. The Neolithic way of life required the introduction of agriculture was fol- other sorts of planning than the Meso- lowed by this first monumental architec- lithic. A given area could produce more ture in the shape of megalithic tombs food, and the population density must with stone-built chambers and heavy kerb have increased dramatically during the stones. However, the megalithic tombs 4th Millennium BC. With the Neolithiza- were not built immediately after the pri- tion the societies became more seden- mary Neolithization. A couple of hun- tary, and much work and planning were dred years, at least, should be allowed invested into a certain territory. An infor an establishment or initial phase of creasing need for marking the territory agriculture before the large megalith was a natural consequence of this de- building activity began. It seems to be velopment. Important places in the opened close to a rule of thumb that 300-500 landscape became focus for the first years should elapse before the Stone 101

Age societies reached this point of mon- an intensive and systematic recording umental activity. In Denmark the build- programme of all prehistoric monuments, ing of dolmens began around 3500 BC, initiated in 1873, and directed from the while the larger passage tombs were National Museum, we have knowledge constructed at 3300-3100 BC (Jensen of many megalithic tombs that have dis- 2001: 359 ff.; Dehn & Hansen 2006). The appeared since then: Altogether c. 7000 stone-built tombs do not represent the megalithic tombs have been registered earliest monumental burial architecture. (Ebbesen 1985: 12; Dehn, Hansen & Kaul During the preceding centuries earthen 2000: 7). Earlier recordings of tombs not long barrows were built, often with registered during the large campaigns monumental timber facades (Kaul 1988; of the National Museum hint at a much Kjær Kristensen 1991; Andersen & Johan higher numbers of megalithic tombs than sen 1992). However, there still seem to 7000. Many dolmens and passage tombs be a gap of a couple of centuries after must have been destroyed already dur- 4000 BC, before this building activity ing the medieval period giving building commenced. The same pattern seems to material for churches; other tombs must emerge in England. Here the general as- have served as quarries for building of sumption has tended to be that monu- castles and manor houses of the 16th and mental architecture – long barrows and 17th century. Consequently, it seems to stone chambered tombs – was intro- be a sound estimate that around 25000 duced at the very beginning of the Neo- megalithic tombs were built in what is lithic. However, among others a recent now Denmark (about 43000 square kilo 14C-dating programme has yielded results meters). About a tenth of what was orig- that make it unlikely that long barrows inally built has survived till today (Eb- and long cairns appeared in southern besen 1985: 37-40; Kaul & Krogh 1990: Britain before 3750 BC (Thorpe 2009: 30). 245-257; Dehn, Hansen & Kaul 2000:7; The building of smaller megalithic tombs Ebbesen 2007: 30-33). in western Britain, such as the portal dolmens, may have begun a bit earlier. When considering the time brackets for the building period of the megalithic Denmark should be highlighted as the tombs – 3500 BC-3200/3100 BC – a rel- country or region in Europe with the atively short period of 300-400 years, highest density of megalithic tombs. To- then the megalithic achievement seems day around 2400 dolmens and passage to be even greater. In average about 150 graves are – more or less – preserved, megalithic tombs should have been and under national guardianship. Due to completed every year, and considering a

Nordlige Verdener Shetlandsprojekt 102

certain building period especially for the and in Spain and Portugal the monu- larger passage tombs, then in every ment density is as high as in Denmark. ‘parish’ one would have seen building That goes also for Orkney, but even work in progress. Shetland with the northernmost megalithic tombs, including around 55 record- Compared with other megalithic regions, ed sites (Henshall 1963), should be anti such as Ireland, the recorded number of cipated as a ‘high density area’. And of tombs of Denmark – c. 7000 – is quite course, as in South Scandinavia, many high. In Ireland ‘only’ c. 1450 megalithic more monuments must originally have tombs have been recorded (Shee Twohig been constructed. 1990: 8). On the other hand, Denmark (and South Scandinavia) can not present The northernmost megalithic enormous labor consuming monuments tombs of Scandinavia or monument complexes such as those Denmark, South Sweden (Scania) and from Ireland. Scotland can boast of around parts of North Germany represent the core 600 megalithic tombs, with a much var- areas of the megalith building achieve- ied density. In certain areas of Brittany, ment of the people of the Funnel Beaker

Fig. 1: The Skjeltorp dolmen, Øst- fold, South Norway, re- constructed F. Kaul Photo. and removed some meters from its original location. 103

F. Kaul Photo.

Fig. 2: Culture. In South-western Scandinavia, bers amber beads of types related to the The northernmost passage the Neolithic expansion of the Funnel Funnel Beaker culture have been found, tomb of Beaker Culture halted close to what is and a 14C-sample has yielded a date to Scandinavia, now the Swedish-Norwegian border at around 3500 BC (Østmo 2007; Glørstad Massleberg, Skee Parish, Svinesund. Just north of this border, on 2009). On the eastern side of the Oslo Bohuslän, each side of the Oslo Fjord, we find the Fjord, a bit further south, at Skjeltorp in Sweden. northernmost dolmens of Continental Østfold, there is a partly preserved dol- Europe. At Holtnes on the Hurum Penin- men, with finds of Funnel Beaker pottery sula at the western side of the Oslo Fjord (Fig. 1). It should be noted that the meg- there are two small dolmen chambers alithic tombs of Shetland are situated where some of the kerb stones of the higher North than the northernmost ones mound are preserved. In one of the cham- from Norway (Kaul 2011: 46).

Nordlige Verdener Shetlandsprojekt 104 F. Kaul Photo.

Fig. 3: The passage tomb, Sned- stenan, Tanum Parish, Bohuslän, Sweden. One of the capstones, now tilted, is covered with later Bronze Age cup marks.

These few South Norwegian dolmens Västergötland, within an inland triangle are relatively small, and even though of fertile land of c. 50 by 30 km there is some more megalithic tombs must have a remarkable concentration of large pas- been built, the building activity here at sage tombs (Fig. 4). Around 300 mega- the border of the Neolithic expansion lithic tombs are recorded here, most of was seemingly not overwhelming. But them being passage tombs (Persson & as soon as we go further south, crossing Sjögren 2001: 6). From Gothenburg there what is now the Norwegian-Swedish is not far along the Halland coast before border, moving into Bohuslän, we find reaching Scania and the core area of more megalithic tombs including a number megalithic tombs. of large passage graves. At Massleberg, just south of the border, a typical pas- Multi-period construction sage grave should be highlighted as be- in Denmark ing the northernmost of this type of Even though many or most of the mega- monument (Fig. 2). Even larger is the pas lithic tombs seem to have been con- sage tomb Snedstenan in Tanum Parish structed within one building phase, as one (Fig. 3). Further south, in Fallbygden, single lay-out, in particular the larger 105

monumental passage tombs, a smaller megalithic kerb stones, and a passage number of monuments are characterized grave chamber. Outside these kerb- by showing distinct separate building stones the excavation revealed a row of phases. free standing stones (Rønne 1979). An- other example is a long dolmen from Firstly, there are earthen long barrows, Troelstrup in northern Central Jutland. It which at a later time became incorpo- all began with a long barrow with wood- rated in and covered by a larger mega- en revetment and a half wooden, half lithic structure. A good example is the stone-built chamber. In the following se- long barrow from Bygholm Nørremark, quence the long barrow was extended with wooden burial structures and other in order cover succeeding chambers, mortuary structures, themselves demon both megalithic and ‘half-megalithic’. strating a complicated sequence – later The last two chambers are true mega- covered by a large long barrow with huge lithic dolmens chambers, the one being

Fig. 4: Large passage tomb, Karleby, Fallbygden, Västergötland, Sweden.

F. Kaul Photo.

Nordlige Verdener Shetlandsprojekt 106

Fig. 5: Multi-period construction of Danish Dolmens; upper: Dolmen from Frejlev, Lolland, Denmark; lower: Dolmen from Lerbjerg, Central Zea- land, Denmark. After A. P. Madsen 1896).

rectangular the other polygonal. Finally In a number of cases traces of multi-pe- the enlarged long barrow was framed riod construction of megalithic tombs by large megalithic kerb stones (Kjærum themselves can be observed just by con- 1977). sidering the immediately visible remains. 107 F. Kaul Photo.

Fig. 6: Dæmpegårds dyssen or Kongedyssen, Tokkekøb Hegn, West of Copenhagen, Zealand, Den- mark,

At Gundsømagle, North Zealand, and at The finest Danish example of a mega- Frejlev, Lolland, the primary structure lithic tomb showing multi-period-con- was a dolmen chamber with its small struction comes from Lønt, South Jut- barrow surrounded by an oval setting of land, where a totally ruined tomb was megalithic kerb stones, subsequently excavated during 1987. The excavations being enlarged by a rectangular setting revealed a long building sequence of of kerb stones making it into a long dol- four separate construction phases, each men (Madsen 1896: Pl. XVI). Another ex- reflected by its own megalithic tomb ample is a dolmen from Lerbjerg at Hval- (Jørgensen 1988). When the monument sø, central Zealand, with a rectangular was complete it was a 40 meters long inner kerb, the barrow subsequently en- and 8 meters wide long barrow sur- larged into a long dolmen (Fig. 5). Dæm- rounded by large kerb stones. The oldest pegårdsdyssen west of Copenhagen phase, structure I, was a round dolmen provides yet another example. Here the surrounded by large megalithic kerb primary structure is a rectangular dol- stones. In the middle of the barrow there men chamber surrounded by a small was a small square stone chamber with- oval mound with small kerb stones. A out a passage. Immediately west of it larger rectangular dolmen chamber was lay a further round dolmen, structure II. added and the fist oval mound became The mound was similar to the first one, incorporated in a large long dolmen with but the chamber had a short passage. huge kerb stones (Fig. 6). Structure IV and structure III, both pas-

Nordlige Verdener Shetlandsprojekt 108

sage tombs, were placed in mound ex- British Isles (Smith & Lynch 1987: 31). tensions to the east and to the west of Often a small round cairn with one rela- the earlier dolmens, and a full long tively simple megalithic chamber com- mound was being created (Fig. 7). prises the primary monument. After a certain interval of time the monument Multi-period construction was enlarged, and the original round in Britain barrow became incorporated in a long From Britain, in particular from Wales and cairn, often with a distinct facade at the Scotland, excavations have revealed ev- opening of a chamber. idence of multi-period megalithic tombs from a larger number of sites. Such com- This is the case at the megalithic tomb posite tombs are now established as of Trefignath on the isle of Anglesey, being a widespread phenomenon, and North Wales, where the first phase con- examples are known throughout the sists of a small round cairn (Fig. 8). The 109

Fig. 7: Plan drawing of the multi-period megalithic tomb from Lønt, South Jutland. Incorporated in one megalithic long barrow we can follow the development of the megalithic tombs, from the oldest, closed rectangular dolmen chamber to the fully developed passage tombs. The hatched areas represent clay packing around the chambers. After Jør- gensen 1988.

the cairn was extended eastwards so as to incorporate the new chamber. The former front and forecourt was obscured by the enlargement of the cairn and a new revetment wall marked its eastern part (Smith & Lynch 1987: 10 ff.).

Fig. 8: A somehow similar arrangement is seen Fig 9: The sequence of the at Dyffryn Ardudwy in southern Wales, Mid Gleniron megalithic tomb of I, Wigtown- Treffignath, Anglesey. where the primary chamber was a por- shire, Scot- Wales. After Smith tal dolmen. This chamber originally pos- land. Two & Lynch 1987. sessed a small round/oval cairn. A much small cairn with a mega- larger two compartment chamber was lithic cham- after a certain period constructed east ber became of the first chamber, and the original incorporated in a long chamber and cairn became incorpo- cairn. After rated in a long, slightly wedge- Corcoran shaped cairn (Powell 1963; 1972. Lynch 1969: 133-135).

The same sequence of multi- next phase saw the addition of a second period construction has been chamber and re-planning of the cairn, observed in Scotland where now a long cairn of wedge shape. The the evidence from Mid Glen ‘new’ chamber opened in a deep, almost iron, Wigtownshire, South- funnel-shaped forecourt/facade marked west Scotland, is of partic- by revetment walls of dry-stone walling. ular interest (Fig. 9). Mid After a further interval a third chamber Gleniron I comprised two was constructed to the east, partly within small oval/round cairns arranged in the forecourt of the second chamber and tandem, each containing a rectangular blocking access to it. At the same time chamber, which had subsequently been

Nordlige Verdener Shetlandsprojekt 110

incorporated in a short straight-sided long cairn with a megalithic concave façade at the northern end. A third chamber, set between the oval cairns, appears to have been built contempora- neously with the long cairn (Corcoran 1972: 36). Here, apparently, a concave façade is an addition to a chamber in a round cairn.

A similar arrangement, though only with one chamber, has been observed much Fig. 10: Balvraid, Inverness-shire, west coast, further north along the west coast of Scotland. A small round cairn has been incorporated in a larger cairn with a façade. After Corc- Scotland, in the cairn at Balvraid near oran 1972. Glenelg, West Inverness-shire (Fig. 10). The original structure was a passage tomb, enclosed in a circular cairn, to which an almost square cairn with a façade making this shape (Fig. 11). Cor- slightly concave façade had been added coran expresses himself as follows: (Corcoran 1972: 34-35). “There is nothing to suggest, however, that the heel-shaped structure at Tulach Caithness, multi-period an t-Sionnaich was not added to a circu- construction lar cairn after the latter had already en- One of the most interesting megalithic joyed an independent, although possibly tombs showing multi-period construc- short, existence. Without the heel- tion is the long cairn of Tulach an t-Sion- shaped structure the latter is a simple naich at the northern Caithness coast, Passage Grave, set in a circular cairn…” excavated during 1961 (Corcoran 1966: (Corcoran 1966: 16). Naturally it should 1-22; Corcoran 1972: 32-34; Henshall be a matter of debate how much time 1991:146-149). The primary structure has elapsed between building the round was a round cairn with a rectangular cairn and the subsequent addition of the passage grave chamber to which a heel- heel-shaped cairn with its concave shaped cairn with a concave façade had façade – 10 years, 50 years, 100 years, been added. The term ‘heel-shaped’ re- or more? – The multi-period construc- fers to the shape of a shoe’s heel, the tion of the heel shaped cairn at Tulach oval/roundish cairn with its concave an t-Sionnaich might even consist of 111

Fig. 11: Tulach an t-Sionnaich, Caitnhess, Scotland. Excavation plan of the heel-shaped cairn. After Corcoran 1966.

more elements, since there were found 10; 1972: 32). But could it be totally ex- parts of both an inner and an outer re- cluded, that we are dealing with an in- vetment wall of the round cairn. The independent building phase, where this ner wall has been regarded as a part of “inner revetment wall”, also including the construction proper, intended to ab- some larger kerb stones, for a certain sorb some of the thrust from the appar- span of time marked the outer limits of ently corbelled chamber (Corcoran 1966: the monument?

Nordlige Verdener Shetlandsprojekt 112

The heel shaped cairn with its concave and even though some robbing of stone façade did not represent the end of the material has taken place, the gap is con- building sequence. Subsequently a large sidered as a possible original feature long cairn was added to the heel-shaped (Davidson & Henshall 1991: 140). This cairn, with a slightly different orienta- might indicate that the long cairn is a tion (Fig. 12). The long cairn in its final later addition. At Warehouse South, not appearance then looks like a typical far away from South Yarrows, a steep- Caithness long cairn. sided mound with a complex passage grave chamber constitutes the north- The Cairn Tulloch of Assery A not far eastern part of the monument. A rectan- from Tulach an t-Sionnaich provides an gular long cairn extends 37 m southwest example of a round cairn with a passage of the chamber. By an early investiga- grave chamber that has been incorpo- tion of the monument remains of a dou- rated into a cairn of partly heel-shape, ble wall were found, which seem to or a horn-shaped cairn, here even a dou- have surrounded the round mound. This ble horn-shaped cairn (Corcoran 1966; may indicate that the long cairn was a 1972). later addition. A similar situation seems to be represented at Na Tri Shean, where Some of the long cairns of Caithness a wall face is visible at the edge of the should hide similar building sequences: round cairn element, also indicating that The primary round cairns with their the long cairn is a later addition (Ibid.: chambers often being higher than their 154-155 & 128-129). long cairn extensions, and the concave facades or horns with platform arrange- Another arrangement is seen at Cam- ments belonging to subsequent building ster Long where two round cairns origi- phases. The great majority of the long nally covered two passage grave cham- cairns seem to be composite structures bers. Later these two round cairns were (Davidson & Henshall 1991: 47 & 55-59). incorporated in a horned long cairn with ‘false’ façades. When the two round As an example the long cairn of South cairns with their chambers became en- Yarrows North could be mentioned. The closed in the long cairn, access to the monument appears as two distinct ele- chambers was still required, and the ments, in the east a pear-shaped cairn chamber passages were consequently with a complex passage grave chamber, extended to the line of the new outer re- to the west a long cairn extension. Be- vetment (Corcoran 1972: 43; Davidson & tween these two elements a gap is visible, Henshall 1991: 96-102). 113

Among the remarkably few horned long cairns in Orkney the monument at Head of Work, Mainland, seem to provide possible evidence of multi-period construction similar to what has been observed in Caithness. Towards the south- east end the long cairn rises into a prominent steep-sided slightly oval cairn. In various places short lengths of wall-face are exposed in the sloping surface of the round cairn element (Dav- idson & Henshall 1989: 118-119), seemingly forming its original revetment wall, later being included by the long cairn extension.

Shetland Shetland can boast of more than 50 megalithic tombs. They represent the northernmost expansion of the Neolithic monumental architecture. Typologically they form a particular group of megalithic tombs, even though some features show similarities with particularly the Fig. 12: Tulach an t-Sionnaich, Caithness, Scotland. Schematic megalithic tombs of Caithness. The de- drawing of the full multi-period sequence including the long cairn extension. After Corcoran 1972. velopment of the megalithic tombs of Orkney followed other paths. The majority of Shetland’s chambered cairns are façade. However, there is a great varia- described as heel-shaped (Turner 1998: tion among the Shetland megalithic 42), the oval/roundish cairn with its con- tombs, and very few, or perhaps only cave façade making this shape. Some of one, Vementry, combines these features the chambers are of trefoil shape. Alto- in one monument. The chambers are ei- gether this means that a ‘typical’ Shet- ther trefoil-shaped or rectangular in plan land megalithic tomb should be a pas- (Henshall 1963: 139-141), but also some sage grave with a trefoil-shaped chamber polygonal dolmen-like chambers occur. in a heel-shaped cairn with a concave Still, the trefoil-shaped chamber seems

Nordlige Verdener Shetlandsprojekt 114

to be a Shetland peculiarity. Most of the The first phase is represented by a round cairns are – as mentioned – heel-shaped, cairn of 6.3 metres in diameter, enclos- but there are also some round cairns, ing an approximately 2.3 x 2 metres large and even almost square cairns are rep- trefoil shaped chamber, out of which a 2 resented (Calder 1963: 37-40; Turner metres long passage leads south-east. 1998: 43). Subsequently more cairn material was added, particularly to the south-east, One of the most well-preserved passage where a concave orthostatic façade was graves is located on the Isle of Vementry. erected. Also a lengthening of the pas- It is evident, that this monument is of sage must have taken place, even multi-period construction, when observ- though the concave façade blocks the ing the outline and section plan meas- passage opening (Henshall 1963: 177- ured by Calder in RCAHMS 1946 (Fig. 13). 178; Fojut 2006: 25; Mahler 2011: 12-14).

Other larger cairns might have been built in two stages (Fojut 2006: 25). The heel-shaped cairn on Ness of Nouns brough (Henshall 1963: 171) seems to provide a situation similar to that of Ve- mentry. In the cairn there is an inner row of kerb stones, probably marking the pe- rimeter of the primary round cairn. The outer kerb, probably reflecting an enlargement of the cairn, is concave on the south-east landward side, thus defining the heel-shaped cairn with its typical façade (Fig. 14). The heel-shaped cairn at Punds Water may hide an inner primary round cairn (Fig 15). Some visible wall faces inside the heel-shaped cairn might indicate a multi-period construction of this monument, originally

Fig. 13: The Vementry cairn, Shetland, with its multi-period construction. The additional façade marked with red. After RCAHMS 1946, Vol. III. 115

Fig. 14: Plan drawing of the Ness of Nounsbrough heel shaped cairn with its two rows of kerb stones, the chamber area obscured by a wall of a modern watch tower. After Henshall 1963.

from Vementry does not exclude that most of the Shetland monuments were actually constructed in one flow of work. It seems unlikely that for instance the small heel-shaped cairn at Islesburgh should incorporate more building phases (Calder 1963: 45-47).

being a round cairn. It is also suspicious The evidence of multi-period construc- that the long passage changes its orien- tion of the Vementry passage tomb has tation a bit. On the other hand such a not remained unnoticed, including the wall face could be seen as an (inner) striking similarities with the shape and constructional part belonging to one sequence of the Tulach an t-Sionnaich building sequence solely. The evidence heel-shaped cairn, Caithness (Corcoran F. Kaul Photo. Fig. 15: The heel- shaped cairn at Punds Water, Shetland.

Nordlige Verdener Shetlandsprojekt 116

1966: 15-16; 1972: 32-34). At both sites Even though there must have been con- a circular cairn was incorporated in a nections with Orkney as shown by the heel-shaped cairn. Also the sequence domestic architecture, the similarities and shape of the cairn at Balvraid, Inver- with the especially the Caithness mega- ness-shire, shows similarities with Shet- lithic monuments should be noted. Heel- land (Corcoran 1972: 34-35). shaped cairns should not be regarded as being unique for Shetland, and a number Discussion of Caithness candidates could be high- When dealing with multi-period con- lighted. When they are a component of struction and cairn morphology there is long cairns it is difficult to detect the one dissimilarity of great importance heel-shaped cairns (Davidson & Hen- between Caithness and Shetland. As shall 1991: 41-42). More heel-shaped we have seen, the final phase of for in- cairns should hide underneath the high- stance Tulach an t-Sionnaich includes er ‘round-cairn’ elements of the long the addition of a long cairn. No-where in cairns. The evidence from the meticu- Shetland an addition of a long cairn has lous excavation of Tulach an t-Sionnaich been observed, and probably no long should be regarded as representative cairn like the Caithness cairns was ever rather than unique. built in Shetland. There are three cairns, which may have been built in the style When considering that there must have of Orcadian cairns, that is with the been close bonds between Shetland and chambers being divided into compart- Caithness as to the development of ments, formed by slabs on end (stalled megalithic architecture including multi- cairns) (Turner 1998: 43). But possible period construction of similar sequenc- remains of a stalled chamber do not es (see above), the missing long cairns guarantee the presence of a long cairn. in Shetland is a conspicuous fact that The stalled chamber at Houstry cairn, deserves attempts of explanation. It Caithness, is placed in a roundish cairn seems that Shetland became rather iso- with traces of a straight to slightly con- lated in the Bronze Age (Turner 1998: 51 cave façade, the monument close to be- ff.), but also during the Neolithic contact ing a heel-shaped cairn (Davidson & was seemingly highly restricted, and Henshall 1991: 117-118). At Warehouse only very few objects of ‘foreign’ lithic South, Caithness, there is a central, al- material reached Shetland (Ballin 2011). most stalled chamber in a round cairn, The absence of long cairns in Shetland where a long cairn seems to be a later could indicate the time when these dis- addition (ibid.: 154-155). tant isles became more isolated, that is 117

just before the time when the Caithness part or climax of the Neolithization proc- megalithic tombs became enhanced ess. Then, accordingly, many megalithic with their long cairn extensions. It tombs were erected in Shetland around should not be excluded however, that 3300, and perhaps the typologically lat- we are simply dealing with a deliberate est tombs were built one or a couple of choice by the Shetlanders. They did not hundred years later. The peak of mega- create long cairn extensions due to the lithic activity might have been earlier if setting of their Shetland cairns in the the Neolithic began here a bit earlier. landscape, on distinctive knolls and hill- tops, where a long cairn addition, lower At Sumburgh Airport a stone cist has and behind the land mark, would not been excavated containing the disartic- yield any further architectural zest. On ulated bones of at least 18 individuals. the other hand, some monuments like 14C dates demonstrate that the cham- the heel-shaped cairn at Islesburgh ber was in use about 3300 BC (Turner would architecturally benefit by a long 1998: 41; Mahler 2011: 12). The burial cairn addition up the hill. structure is described as “an unusually large cist” made by boulders set on When: concluding remarks edge, which may originally have been The evidence from the excavations of covered by a mound (Turner 1998: 41). the kitchen midden at West Voe near When using the word ‘cist’, a Late Neo- Sumburgh could indicate Neolithic set- lithic or Bronze Age dating might be tlement in Shetland around 4000 BC hinted at. However, why not consider (Ballin 2011: 33; compare Gilmore & this tomb as a megalithic tomb proper, Melton 2011: 69 f.; Melton 2009: 184). belonging to the ‘mainstream’ of mega- The evidence is sparse for Early Neolith- lithic tradition? It could very well be ic settlement on Shetland, and it has classified as a rectangular dolmen-like proposed that the introduction of agrar- chamber – and – within a mound. ian economy first took place at 3700 or 3600 BC (See Alison Sheridan, this vol- When returning to the Scottish main- ume, with further references). When its land in general, farming was present at seems to be close to a rule of thumb some stage soon after 4000 BC (Ash- that 300-500 years should elapse from more 1996; Warren 2004: 96), while a the first neolithization until the Stone statistical reassessment of the 14C Age societies reached the point of meg- dates may indicate a major shift in the alithic building activity, the creation of economy around 3800 BC (Ashmore stone monuments being an integrated 2004: 125 ff.). Available 14C dates of

Nordlige Verdener Shetlandsprojekt 118

charred barley are concentrated in the Tulach an t-Sionnaich; that is the time archipelagos and in eastern and south- around 3500 BC (Davidson & Henshall eastern Scotland, and they suggest that 1991: 83-84; Ashmore 1996: 29-33). barley was grown from about 3700 BC. It seems difficult to say whether there Most of the dates obtained from cattle was a delay of the Neolitization of Shet- bones come from Orkney, and they seem land as to mainland Scotland. Since to show that the practice of depositing some of the megalithic tombs of Shet- these bones at settlements and other land seem related to these of Caithness sites started about 3500 BC (ibid.: 127- they may have been built within the 129). The large timber house or hall from same period of time. What we need is Balbridie, Aberdeenshire as well as a more precise dates for the long cairn ex- similar construction from Claish Farm, tensions of Caithness. Then we would Stirling, may have been built in the first perhaps be able to give a clearer esti- or second quarter of the 4th Millenium mate of when the Shetlanders ‘jumped BC. 14C dates from the stone-built hous- off’ the development. Would the years es of Knap of Howar, Papa Westray, Ork- around 3400 BC be a compromise for ney, suggest that they were constructed the time being, or perhaps a century lat- about 3500 BC (ibid.: 133). er? Was it from then that Shetland became increasingly isolated, when not in- When the statistical ‘weakness’ of the troducing long cairns? Other questions 14C dates is considered there is no evi- remain. Would it be possible to include dence that chambered tombs were built other evidence of contacts or lack of in Scotland before about 3700 BC. A hu- contacts in order to confirm such possi- man bone from Tulloch of Assery A, ble dates, for instance the presence or Caithness, has yielded a date between absence of exotic lithics in Shetland. 3950 and 3300 BC, and an animal bone As mentioned above the possibility re- from at Tulloch of Assery B has yielded a mains that the absence of long cairns date between 3990 and 3520 (Ashmore was not due to isolation but as a delib- 2004: 130). These dates are from cham- erate choice by the Shetlanders. How- ber deposits, and they do not date the ever, it is the hope of the present author construction of the tomb, but the use of that this analysis of megalith typology the tomb, even though they could mark and the evidence of multi-period con- an early use of the tomb. There are oth- struction could make a small contribu- er 14C dates from Scottish megalithic tion to the discussion of the cultural tombs that suggest the same time for processes and connections of the Neo- (early) use of the chambers, including lithic societies at the border. 119

Bibliography Corcoran, J.X.W.P., 1966: The Excavation of Three Chambered Cairns at Loch Cal- Andersen, S.H. & Johansen, E., 1992: der, Caithness. Proceedings of the Society An Early Neolithic Grave at Bjørnsholm, of Antiquaries of Scotland XCVIII: 1-75. North Jutland. Journal of Danish Ar- chaeology 9, 1990 (1992): 38-58. Corcoran, J.X.W.P., 1972: Multi-Period Construction and the Origins of the Ashmore, P.J., 1996: Neolithic and Chambered Long Cairns in Western Brit- Bronze Age Scotland, Historic Scotland, ain and Ireland. Prehistoric Man in Wales London. and the West, essays in honour of Lily F. Chitty. Lynch, F. & Burgess, C. (eds.), Ad- Ashmore, P.J., 2004: Absolute chronology. ams & Dart, Bath, Somerset: 31-63. Scotland in Ancient Europe. The Neolithic and Early Bronze Age of Scotland in their Davidson, J.L. & Hanshall, A.S., 1989: European Context. Shepherd, I.A.G. & The Chambered Cairns of Orkney, Edin- Barclay, G.J. (eds.), Society of Antiquar- burgh University Press, Edinburgh. ies of Scotland, Edinburgh: 125-136. Davidson, J.L. & Hanshall, A.S., 1991: Ballin, T.B., 2011: The Post-Glacial Colo- The Chambered Cairns of Caithness, Edin nization of Shetland – Integration or Iso- burgh University Press, Edinburgh. lation? Evidence from Lithic and Stone Assemblages. Farming on the edge: Cul- Dehn, T. & Hansen, S.I., 2006: Birch bark tural Landscapes of the North, 2011: in Danish passage graves. Journal of Short papers from the network meeting Danish Archaeology 14, 2006: 23-44. in Lerwick, Shetland September 7th-10th 2010. Mahler, D.L. & Andersen, C. (eds.), Dehn, T, Hansen, S.I. & Kaul, F., 2000: Northern Worlds, the National Museum Klekkendehøj og Jordehøj. Restaurerin of Denmark, Copenhagen: 32-43. ger og undersøgelser 1985-90, Stenalder grave i Danmark Bind 2, Nationalmuseet, Calder, C.S.T., 1963: Cairns, Neolithic Skov- og Naturstyrelsen, Copenhagen. houses and burnt mounds in Shetland. Proceedings of the Society of Antiquar- Ebbesen, K., 1985: Fortidsminderegistre ies of Scotland XCVI: 37-86. ring i Danmark, Fredningsstyrelsen, Copen hagen.

Nordlige Verdener Shetlandsprojekt 120

Ebbesen, K., 2007: Danske dysser. Dan- Kaul, F., 1988: Neolitiske gravanlæg på ish Dolmens, Attika, Copenhagen. Onsved Mark, Horns Herred, Sjælland. Aarbøger for Nordisk Oldkyndighed og Fojut, N., 2006: A guide to Prehistoric Historie 1987 (1988): 27-83. and Viking Shetland, The Shetland Times Ltd., Lerwick. Kaul, F., 2011: Early Agriculture at the Border. Farming on the edge: Cultural Gillmore, G. K. & Melton, N., 2011: Early Landscapes of the North, 2011: Short Neolithic Sands at West Voe, Shetland papers from the network meeting in Ler- Islands: Implications for human settle- wick, Shetland September 7th-10th 2010. ment, Geological Society, London, Spe- Mahler, D.L. & Andersen, C. (eds.), cial Publications 2011; v. 352: 69-83. Northern Worlds, the National Museum of Denmark, Copenhagen: 44-57. Glørstad, H., 2009: The Northern Prov- ince? The Neolithisation of Southern Kaul, F. & Krogh, K. J., 1990: En lidet Norway. Neolithisation as if history påagtet stenkløvningsteknik – og om mattered. Glørstad, H. and Prescott, C. oldtidsminder som stenbrud. Aarbøger (eds.), Bricoleur Press, Lindome: 135- for Nordisk Oldkyndighed og Historie 168. 1990: 225-260.

Hellgren, M. & Johansson, L., 2007: Kjær Kristensen, I., 1991: Storgård IV. Fornminnen i Bohuslän. Från mellersta An Early Neolithic Long Barrow near och norra Bohuslän. Fjelsø, North Jutland. Journal of Danish Archaeology 8, 1989 (1991): 72-87. Henshall, A.S., 1963: The Chambered Tombs of Scotland, vol. one, Edinburgh. Kjærum; P., 1977: En langhøjs tilblivelse. Antikvariske studier 1 (Nationalmuseet): Jensen, J., 2001: Danmarks Oldtid, Ste- 19-26. nalder 13.000-2.000 f.Kr., Gyldendal, Co- penhagen. Lynch, F., 1969: Megalithic Tombs of North Wales. Megalithic Enquiries in Jørgensen, E., 1988: Fire storstensgrave the West of Britian. A Liverpool Sympo- i en højtomt ved Lønt. Nationalmuseets sium. Powell, T.G.E, Corcoran, J.X.V.P, Arbejdsmark 1988. Lynch, F. & Scott, J.G. (eds.), Liverpool University Press, Liverpool: 107-148. 121

Madsen, A.P., 1896: Gravhøje og Grav- Shee Twohig, E., 1990: Irish Megalithic fund fra Stenalderen i Danmark. Det Tombs, Shire Archaeology, Risborough. østlige Danmark, Gyldendalske Boghan- del, Copenhagen. Smith, C.A. & Lynch, F.M., 1987: Trefig- nath and Din Dryful. The Excavation of Melton, N. D. 2009: Shells, seals and Two Megalithic Tombs in Anglesey, ceramics: an evaluation of a midden at Cambrian Archeological Monographs West Voe, Sumburgh, Shetland, 2004-5, No 3, The Cambrian Archaeological As- Mesolithic Horizons, Papers presented sociation, Bangor. at the Seventh International Conference on the Mesolithic in Europe, Belfast Turner, V., 1998: Ancient Shetland, 2005, Vol 1: 184-189, McCartan, S., Amenity Trust, Historic Scotland, Lon- Schulting, R., Warren, G. and Woodman, don. P. (eds.), Oxbow Books. Warren, G., 2004: The start of the Neo- Persson, P. & Sjögren, K.-G., 2001: Fall- lithic in Scotland. Scotland in Ancient bygdens gånggrifter. Undersökningar Europe. The Neolithic and Early Bronze 1985 till 1998. Gotarc Ser. C Nr. 34, Age of Scotland in their European Con- Göteborgs Universitet, Gothenburg. text. Shepherd, I.A.G. & Barclay, G.J. (eds.), Society of Antiquaries of Scot- Powell, T.G.E., 1963: The Chambered land, Edinburgh: 91-102. Cairn at Dyffryn Ardudwy. Antiquity 37: 19-24. Thorpe, N., 2009: Becoming Neolithic in Southern Britain. Neolithisation as if RCAHMS, 1948: Twelth Report with an history mattered. Glørstad, H. and Pres- Inventory of the Ancient Monuments of cott, C. (eds.), Bricoleur Press, Lindome: Orkney & Shetland. Vol III. Inventory of 23-63. Shetland, Edinburgh, His Majesty’s Sta- tionary Office. Østmo, E., 2007: The Northern periphery of the TRB. Acta Archeologica 78, 2: Rønne, P., 1979: Høj over høj. Skalk 1979, 111-142. nr. 5: 3-8.

Nordlige Verdener Shetlandsprojekt 122

The northernmost Bronze Age Farms

Preben Rønne

Introduction Becker from Copenhagen was excavat- The first Bronze Age house remains in ing late Stone Age graves near Ristoft in Southern Scandinavia were excavated western Jutland, and it was necessary in 1955 in Fragtrup Western Jutland – to move the top soil from a large area. Fragtrup House I (Draiby 1985: 127-216). By doing this he uncovered the remains It was found by pure chance by the farm- of three house sites appearing as discol- er, who found the stone built fireplace, oration in the yellow subsoil. The house and it was later excavated by the son of remains, fig. 1, were of the same type as the local teacher. The house was a three at Fragtrup and almost of the same size, aisled longhouse, 18 m long and 7 m dated to the Late Bronze Age Period IV wide and relatively well preserved with (Becker 1968: 79-88). Between 1968 and a fireplace, clay floor and culture layers. 1972 another six house remains were House II was excavated some years lat- excavated at Hovergård, and not far er by the National Museum of Denmark from Hovergård at the village of Spjald during 1957. The Bronze Age farm hous- more than 40 house remains were un- es from Fragtrup were not published uncovered together with 30 found at Bjerg til 1985 maybe because they differed some kilometres north of Spjald all in from what was expected of the Bronze western Jutland and from the Late Age architecture during the 1950’ies. Bronze Age (Becker 1972). (Draiby 1985: 127-216) The result of these large scale excava- During 1967 new finds made it generally tions using a mechanical excavator to accepted that Bronze Age houses were strip off the top soil, more than 80 house well built, three aisled and with strong remains were found within a few years, roof-bearing posts. Many times in ar- all from the late Bronze Age, per V-VI, c. chaeology the new finds were uncov- 900-500 BC and all were built by the ered by pure chance. Professor C.J. same pattern. 123

Norway – the first remains of farm houses The Danish large scale excavations from the 1960’ies and early 1970’ies were of course known in Norway, but only a few Norwegian archaeologists tried to copy it. It was not until a research project on Forsand in Rogaland, using a mechanical excavator to strip off the top soil the method was more generally introduced in Norway. The method became more common from the middle of the 1990’ies and since then used regularly. Then – of course – Bronze Age houses were found. During the years 1980 to 1990, 1992 and 1994 about 80.000 m2 were examined at Forsand and approximately 250 house remains were recognised whereof 20 were dated to the Bronze Age (Løken 1998: 170–171).

The houses in Rogaland were all of the same type and construction as the South Scandinavian house remains. Within a few years more than 300 house remains from the Bronze Age and Early Iron Age were discovered. The development of house types in Rogaland is shown in fig. 2.

Fig.1 (top right): Late Bronze Age house from Ristoft, Western Jutland. The remains indicate a house measuring 19,5 m by 6 m. Courtesy Becker 1968.

Fig. 2: House types from Rogaland from Late Neo- lithic to Pre-Roman Iron Age. Courtesy Østmo, Hedeager 2005:186.

Nordlige Verdener Shetlandsprojekt 124

Northern Norway

The northernmost Bronze Age Farms One of the most spectacular finds is the Bronze Age house from Kveøy, Troms County (fig. 3). It was excavated in 2008 and 2009 (Arntzen and Sommerseth eds. 2010). The must have been longer than house was partly disturbed 12 m and might have been c. by a road, and only traces 5–7 m wide (fig. 4). of six post holes were left. It was not possible to estab- The house remain is dated by a radio- lish the length or how broad carbon date from charcoal from one of the house structure origi- the postholes, and gives the date be- nally had been, and no tween 900-770 BC or the Late Bronze traces of the walls had Age. The construction is just like the survived, but it is es- Bronze Age houses further south, which timated that the clearly shows the connections with the house originally traditional Nordic Bronze Age societies from Rogaland and South Scandinavia. Furthermore, the nearest house remains dated to the Bronze Age are laying c. 1000 km South of Kveøy near Trondhe- im! Near the house remains at Kveøy an old cultivation layer was excavated, dated to the Late Bronze Age and containing grains of barley and wheat presumably showing what had been cultivated in the area.

Fig. 3: Sites with Bronze Age Fig. 5: Torgårdsletta, Sør-Trøndelag County before house re- the archaeological excavation during 1998. A mains men- large gravel pit has disturbed most of the archae- tioned in ological site, and only a small area has been left the text. for later excavation. Courtesy Solheim 1999. 125

Fig. 4: The northernmost house remains from the Bronze Age, excavated on Kveøy, Troms County. Courtesy Arntzen 2010a.

Central Norway near Trondheim. Only a minor part of the area with Bronze Age house remains Torgårdsletta, Kvenild, have been excavated, as a gravel pit has Sør-Trøndelag County changed the landscape dramatically (fig. The remains of several Bronze Age farms 5) (Solheim 1999: 9 Grønnesby 1999: 10). have been found in Sør-Trøndelag County, The house sites at Kvenild were exca- and one of the most important excava- vated by Vitenskapsmuseet (Museum of tions is fromThorgårdsletta, Kvenild farm Natural History and Archaeology, Trond-

Nordlige Verdener Shetlandsprojekt 126

Fig. 6: Kvenild on Torgårdsletta. Most of the area has been disturbed before the excavation. The postholes have been marked by house remains, and the project manag- white plates. Courtesy Grønnesby1999. er, Geir Grønnesby, assumed that the many traces of postholes indicated that heim) during 1998, and probably only a many more house remains have origi- minor amount of the farm remains have nally been present at the site. Grønnes- survived the digging of the gravel pit, by assumed that at about five household and the “village” must have been much units have existed simultaneously in the larger. Twenty house sites were con- area for every 100 year period (Grønnes- firmed (fig. 6 and 7). The oldest house by 2005: 97). remains have been dated to 1200-100 BC, but most of the remains are from Stokkset, Møre and 1200-100 BC. Romsdal County During the late 1970’ies the first two Until recently the house remains from aisled house remains from late Neolith- Kvenild were the northernmost long ic and Early Bronze Age were excavated houses from the Bronze Age in Norway. in Denmark. But the house type was ac- A large amount of post holes could not tually excavated in Norway as early as be connected to any known type of 1953–55 by Egil Bakka (Løken 1998: 170; 127

Børsheim 2005: 115, 116 fig. 4). Interest- ing enough they were excavated even earlier than the house remains at Fragtrup mentioned above, but because of the question of dating the house remains at Stokkset were not published until the late 1970’ies. Today the house remains dating to the Late Neolithic or Early Bronze Age is generally accepted.

Søberg, Sør-Trøndelag County Two Bronze Age house sites are known from Melhus municipal in Sør-Trøndelag c. 20 km south of Trondheim, Søberg and Skjerdingstad, and both sites are located in the broad valley along the river Gaula, Fig. 7. fig. 8. The earliest house remains at Sø The 20 houses berg are probably connected to the ini- at Kvenild. tial phase of agriculture in the area, and The houses are dated from the site represents the only place in Early Bronze Trøndelag where two aisled houses Age and Pre- have been found. The soil is very rich Roman Iron Age; Mona and well suited for agriculture, and at Ødegården the same time both the river Gaula is Del. one of the best fishing places in Norway and the mountains around the valley are rich in game.

The earliest long houses from late Neo- lithic, c. 2000-1700 BC, and Early Bronze Age Period I, c. 1700-1500 BC, are very

Fig. 8: The landscape along the river Gaula, where Bronze Age house remains have been excavated at Søberg and Skjerdingstad. Photo Fjellanger Widerøe AS.

Nordlige Verdener Shetlandsprojekt 128

E6 Søberg, 2003 Felt III Storsand 90/2, Melhus, Sør-Trøndelag Tegning: M. Ødegården

0 5 10 15 m N

Førromersk Jernalder 500-1 f. Kr. Hus I Bronsealder 1800-500 f. Kr.

BC400-200

BC390-200 BC410-200 BC480-200

BC390-200 Hus IV

Hus VII

BC1010-790 BC1310-1000 Hus V

Hus II BC1050-800

Hus III BC1000-820 BC830-540 BC530-250

grus

BC980-780

BC1120-910

Hus VIII Hus VI Fig. 9: Søberg III with two two- aisled house remains from the BC820-540 Late Neolithic or Early Bronze Age per. I. Also three aisled houses from Bronze Age and Pre-Roman Iron Age can be seen. Mona N Ødegården and Turid Brox Nilsen BC800-540 del.

0 10 20 m 129

10. In front the remains of house VI from Søberg III. It is a typical scarce in Norway, and they are only house from the Bronze Age in Melhus, 14 C dated to 820 – 540 BC. The house has stables in the center. In the background houses from known from tree places (fig. 9). They fol- Bronze Age and Early Pre Roman Iron Age. The two aisled houses low partly the same building technique were found later. Photo Vitenskapsmuseet. as known from southern Norway and from South Scandinavia with one central row of roof bearing posts. In Rogaland the Bronzes and farms house type is known from eight archaeo In Central Norway only one bronze arte- logical sites, compared to the two house fact is known from the Late Neolithic remains of this type known from a site period. Later during the Early Bronze in Østfold (Børsheim 2005: 109-115). Age Period I, nine bronze pieces have The habitation at Søberg III continued been found. They show very clearly that after Period I with the usual three aisled both the central European and in partic- buildings (fig.10), and the latest house ular the South Scandinavian Bronze age remain is dated to 420-200 BC. Later the culture influenced Central Norway (fig. house structures moved closer to the 11). Today we see the area between river, and the latest house is from Late Sør-Trøndelag and Nord-Trøndelag as a Roman Iron Age c. 200-400 AD. border zone for the expansion of the ear-

Nordlige Verdener Shetlandsprojekt 130

liest Nordic Bronze Age culture, and no Conclusion bronzes or two aisled house remains Central Norway has been a natural part have been found north of this area. The of the European Bronze Age culture, and composition of artefact types is more or with a close association to the central less like in South Scandinavia, where lo- Nordic Bronze Age area in South Scandi cal Nordic types dominate, such as the navia as early as from Period I c. 1800 BC. massive axes of type Fårdrup and local variants of flanged axes.

Fig. 11: Find spots of artefacts from Late Neolithic and Early Bronze Age per. I marked with black. The two aisled farm house remains from the same periods are marked with red. 131

Bibliography Grønnesby, G. 1999: Fortidens hus på Kvenild. Spor 1999, nr. 1: 10-11. Arntzen, J. 2010 a.: Hus og konstruktions spor. In Arntzen, J. & I. Sommerseth (ed.) Grønnesby, G. 2005: Fra stolpehull til 2010: Den første gården i Nord-Norge. hushold. Utgravninger av hustomter på Jordbruksbosetting fra bronsealder til jern Kvenild,Trondheim 1998,Høgestøl, Sels- alder på Kveøy. Tromura. Tromsø Muse ing, M. L., Løken, T., Nærøy A. J. og ums rapportserie Nr. 39, Tromsø: 42- 59. Prøsch-Danielsen L., (red.), Konstruksjons spor og byggeskikk – Maskinell flate Arntzen, J. 2010 b.:Arkeologiske spor efter avdekking – metodikk, tolkning og forvalt jordbruk. In Arntzen, J. & I. Sommerseth ning. AmS-Varia 43, Stavanger: 97-107. (ed.) 2010: Den første gården i Nord-Norge. Jordbruksbosetting fra bronse alder til jern Løken, T. 1998: Det forhistoriske huset i alder på Kveøy. Tromura. Tromsø Muse- Rogaland – belyst ved flateavdekkende ums rapportserie Nr. 39, Tromsø: 109-121. Utgravninger, O. Kyhlberg, O., red., Hus och tomt i Norden under förhistorisk tid. Becker, C. J. 1968: Bronzealderhuse i Bebyggelseshistorisk tidskrift nr. 33:169- Midtjylland. Nationalmuseets arbejds- 184. mark 1968, København: 79-88. Rønne, P. 2005: Arkæologiske under- søgelser forud for bygningen af ny E6 Becker, C. J. 1972: Hal og hus i yngre gennem Melhus, Sør-Trøndelag, Høge bronzealder. Nationalmuseets arbejds- støl, Selsing, M. L., Løken, T., Nærøy A. mark 1972, København: 5-16. J. og Prøsch-Danielsen L., red, Konstruk- sjonsspor og byggeskikk – Maskinell Børsheim, R. L. 2005: Toskipede hus I neo flateavdekking – metodikk, tolkning og litikum og eldste bronsealder; Høgestøl, forvaltning. AmS-Varia 43: 87-96. Selsing, M. L., Løken, T., Nærøy A. J. og Prøsch-Danielsen L., (red.), Konstruksjons Rønne, P. Ip.: De tidligste bronzer i Midt- spor og byggeskikk – Maskinell flateav og Nordnorge, Agrarsamfundenes ekspan dekking – metodikk, tolkning og forvalt sion i Norden – Netværksmøde i Nordlige ning. AmS-Varia 43, Stavanger: 109-121. Verdener, Kaul, F. & Sørensen, L., red., The National Museum of Denmark. Draiby, B. 1985: A Late Bronze Age Set- tlement from Fragtrup, Jutland. Aar- Østmo, E., og Hedeager, L. red. 2005: bøger for nordisk Oldkyndighed og His- Norsk Arkeologisk leksikon. Pax Forlag. torie 1984, København: 127-216. Oslo.

Nordlige Verdener Shetlandsprojekt 132

Neolithic Shetland: Peopling an empty area 4000-3000 BC?

Ditlev L. Mahler

Ditlev L. Mahler photo 2011. 133

Introduction possibility of at least Mesolithic visits As introduced elsewhere in this volume to the Shetland Islands during the peri- the Shetland research project aims to od 5000-4000 BC (Edwards et al. 2009: investigate the Neolithic elements on 113; Edwards & Whittington 1998: 5 f.). Shetland and compare them with the At the time of the Neolithic expansion c. South Scandinavian elements (Mahler 4000-3700 BC Shetland hardly contained 2011: 6 f.). Just mentioning some of the a large Mesolithic population if any, so elements they are house sites, agricul- the Neolithic population probably ex- tural implements, ritual behaviour and panded into empty islands in respect of burial customs and there are many com- a human population. This expansion mon elements. Of cause there are also could very well have arrived on Shetland very clear differences such as the very coming from the South West, and prob- early knife tradition with the Shetland ably not directly from Caithness to Ork- Knives going back to may be 3500 BC ney, and from there to Fair Isle, jumping according to the newest 14C dating (SU- further North to Shetland which would ERC-37997; Mahler 2011a, fig. 2). be expected (Sheridan 2009: 92 f.; Row- ley-Conwy IP). As we shall see later that Compared to Southern Scandinavia, the kind of expansion into empty land is not existing conditions on Shetland are rath- uncomplicated and certain conditions er different, because we do not know must be present for securing a success- for sure that Shetland had a Mesolithic ful expansion. population which could turn Neolithic. Alison Sheridan has written about the Why choose Shetland? early dating at the kitchen midden at There are several reasons to choose West Voe, Sumburgh in this volume and Shetland as one part of a comparative the very early dates of 4200-3600 cal BC analysis. First of all, Shetland is consid- (Melton 2009: 184, see also Melton & ered to be the northernmost area of Eu- Nicholson 2007: 99; Melton personal rope where farming was practised as a comments), fig. 1. Even though none of result of the expansion around 4000 BC. the West Voe finds from the bottom lay- After a standstill of about 1000 years on er can tie the area to a Mesolithic envi- the Continental lowlands just south of ronment, other indications show the the Baltic, the farming economies ex- panded to the North. In Scandinavia, the expansion ebbed away at Svinesund on

Fig. 1: West Voe, Sumburgh, Shetland. The kitch- the border of present day Sweden and en midden is in the center of the picture. Norway 58,5° North or just around the

Nordlige Verdener Shetlandsprojekt 134

present day Oslo Fjord area. The Neoli- (Mahler 2011a & b). This brings me to thisation of the British Isles happened deal with the question of demographic as a push from Western and Northern aspects, wondering how island popula- present day France, and this push tions survive their initial colonization. reached its ultimate boundary approximately 60° North on Shetland, and ac- A pioneer society consisting of a 100 cording to past maritime technology fur- persons may well be economical suc- ther expansion was impossible. cessful, but if 98 of them are men and only 2 women the demographic prospec- Shetland is moreover a closed unit, al- tive is not prosperous. Besides popula- though it is scarcely also a closed labo- tion size we talk about three main ele- ratory. One of the tasks set by this re- ments, namely birth rate or fertility, search project will be to clarify possible death rate and sex ratio (Moore 2001: outward relations after the islands were 397; Moore & Moseley 2001: 526; Rob- first colonised by the farming popula- ert-Lamblin 2006: 235 ff.). It is also im- tion. It is evident that the Neolithic cul- portant to remember that even a fairly tures on Shetland were related to the large population of let say 100 persons, rest of the Neolithic traditions in North- males and females in the reproductive West Europe, and it is just as evident age will, after some hundreds of years, that the Neolithic communities devel- have great difficulties finding an accept- oped their own cultural traits with re- able marriage partner avoiding the soci- gards to tools, building traditions, and ety’s definition of incest. These defini- grave monuments. tions could force the society to an exogamous marriage system (Moore Two cases of pioneer societies 2001: 406). Beside there is not a “magic This paper deals with peopling Shetland number” for a population size that en- as there are different challenges to un- sures a band’s viability in a new environ- derstanding the Neolithisation on the ment. mainland on the British Isles and the continent and on islands especially of “There is no such number. Initial size is the size of Shetland and Orkney. As only one factor contributing to the suc- mentioned in the introduction some of cess of a colonizing group. Probably the Neolithic elements if not all are more important is birth rate, and good present in Shetland, which indicate that luck in having a balanced number of there was a well functioning society on male and females born into the band...” Shetland during the Neolithic Period (Moore & Moseley 2001: 528). 135

Generally we speak of six different models of colonization into empty areas, but here I shall only touch upon three of them (Moore 2001: 395-396; Anderson & Gillan 2000: 43 ff.). The first model and much discussed is called Outpost Model or Leap-frog Model, which de- scribe a human population colonizing an empty area without contact to the mother population. Such expansions are very vulnerable and constitute a demographic risk to say at least. This model is much more suited to situations where the expansion goes into already populated ar- Fig. 2: One of six demographic models. (Moore 2001, fig. 5). eas i.e. with an existing mass of genes, which could explain the rapid Neolithic expansion in continental Europe where sister settlements eventually securing the population of the Mesolithic or Erte- this communication with exchange of bølle/Ellerbech Cultures constitute pos- e.g. raw material for making imple- sible mating or marriage partners for the ments, as we shall see later with the expanding Funnel Beaker (TRB) popula- Sarqqaq Culture of Western Greenland tion (Rowley-Conwy ip: 1ff.). The TRB (Grønnow 2004: 66). The last model to expansion is a large and fast push only be mentioned is the Pulse Model or coming to a temporary halt just North of Wave Model, fig. 2. It is used for de- Bohuslän in Western Sweden or around scribing two beachhead and outpost the Oslo Fjord area as earlier mentioned, scenarios, where colonists arrive in suc- may be because the Mesolithic popula- cessive groups securing a steady gene tion density drops drastically further flow. The successive arrives of new North and thus the local mass of genes groups could be caused by special at- also drops. tractive elements in the newly colonized land such as arable land with the possi- The String of Pearls is a variation of the bility of a social rise or special luxury Matrix Model (Moore 2001: 395) and raw material such as walrus teeth, if we describes a colonization along coast- look at Norse Greenland. In these far lines or river systems securing continu- out societies it is necessary for some ous communication with neighbouring kind of social economic relations, which

Nordlige Verdener Shetlandsprojekt 136

the archaeologist could recognise as ex- during the 15th Century for the Norse otic objects and the anthropologist e.g. population in West Greenland (Mahler ritualized objects. These phenomenons 2007: 412 ff.). The cause of this depopu- could indicate a gene flow so to speak lation should probably be seen as part as the most important function. One of of demographic development in Scandi- the most famous examples in the an- navia and Europe as a whole. After mid thropological research is the Kula Ex- 14th Century there were waves of dis- change Ring and Red Feather Money eases not at least the Black Death, both in Oceania (Malinowski 1922; which reduced the European population Neich 2006: 217 ff.; Kirch 1988:103 ff.). with between 30 and 60 % (Lynnerup The Norse expansion in the North Atlan- 1998:122; 2011: 328). There were no tic during the 9th and 10th Century is an pressure on the arable land resources archaeological-historical example. Their any longer, and living on the fringe of dependence on an exchange system the known world far away from family solving the demographic dimension for relations, it must have been much more 500 years, could not prevent the demo- attractive gradually then returning from graphic failure with total depopulation where they had come some 500 years

Fig. 3: Dark band of killiaq in the Østerfjeldet, Nuussuaq, Grønland (Sørensen & Pedersen 2005:110). 137

earlier. In this respect the colonization resource rich and varied Low Arctic area of Greenland was not a failure but a way of West Greenland (Grønnow 2004: 66 of surviving for 500 years. ff.). The Sarqqaq Culture survived for over 2000 years, and the youngest trac- The first paleo Eskimos expansion into es of the culture are dated between 800 the High Arctic started around 2500 BC. and 500 BC. The culture is characterized They crossed the straits between Elles- by bifacial knives, adzes, tanged points, mere Island to the Northernmost Green- micro blades and burins made of Killiaq, land, initiating the Independence 1 Cul- a grey metamorphosed slate (Sørensen ture – part of the Arctic Small Tool & Pedersen 2005: 110; Grønnow & Sø- Tradition. The Independence 1 Culture rensen 2006: 59 ff.), fig. 3. The stone expands to the East through Peary Land quarries for killiaq is situated only in the to areas with a rich population of Musk Disko Bay area, but the killiaq raw ma- Oxen, but North East Greenland is at the terial and artifacts made of killiaq are same time a cul-de-sac from where it is quite common on all the settlements of very difficult to return (Grønnow 2004:78; the Sarqqaq Culture often a very long Grønnow IP.); by 1800 BC the expansion way from the quarries. The exchange was over and the depopulation a fact chains with killiaq could very well be a (Andreasen 2004: 62 f.). The High Arctic mediator for a gene flow along the West is a severe challenge for any human coast of Greenland, and thus creating a population, and even small variations in demographically successful develop- climate and resources may be a threat. ment for the Sarqqaq Culture – for at The demographic aspects could be an least 2000 years (Robert-Lamblin even worse challenge, as the small pop- 2006:235f.). ulations may be very thinly spread over huge distances, so without a steady in- The initial population in Western Oce- flux of new genes the bands were ania arrived some 30.000 year ago doomed to extinction. In fact it is most where they reached as far as the Solo- impressive that the Independence 1 Cul- mon Islands but could not expand fur- ture in fact lasted between 200 and ther because of the lack of maritime max. 700 years. technology. It was impossible to expand out in the eastern Oceania bringing The Sarqqaq Culture, which is the ar- enough provisions at the same time by chaeological name for a contemporary rowing. Only with the invention of the paleo Eskimo culture, did much better sailing technology did it become possi- with a higher population density in the ble for further expansion around 2000-

Nordlige Verdener Shetlandsprojekt 138

1500 BC (Irwin 2006: 62 f.). The expan- Easter Island 900 AD – an expansion sion of the Lapita Culture, a farming covering one third surface of our planet culture, could be due to a mixture of (Anderson 2001:17; Kirch 2007:332; Ir- over population in the already colonized win 2006:67). Some of the Islands be- islands and the wish for social mobility. came depopulated such as Henderson The Lapita expansion is a pulsating ex- Island, Pitcairn, Necker and some of the pansion with periods of stability chang- small Polynesian islands, while other is- ing with renewed pushes and periods of lands showed a considerable population lability. The first push is supposed to growth such as Tikopia (Kirch 2007:334; have happened shortly after 2000 BC, Firth 1959) (fig.). Research into the ex- reaching Hawaii around 700 AD and change networks show connections between as many as 20 or 30 island societies through socio economic transactions.

Fig. 4: Te puke trading canoe. Santa Cruz Islands, about 1900 Kula and Red Feather Money have been (Neich 2006:218). mentioned, and though the systems are fairly recent maybe just two to three hundred years old, they show us the es- sentials of exchange with prestige objects as mediator for a gene flow. -Ar chaeological excavations have shown much older systems among other places on Vanuatu, Vanikoro and Tikopia. Both Kirch and Friedman recognize the exchange importance of the systems for the social reproduction and the forming of marriages especially in the Eastern Lapita area (Kirch 1988: 114; Friedman 1981: 275 ff.).

Conclusion We still need more research before understanding the Neolithic expansion at the ultimate European border around 4000 BC on Shetland. I have already mentioned the possibility that the expansion was a Westerly phenomenon 139

coming up the British West coast, or Fig. 5: was it rather an expansion which used Long totally polished fel- Orkney and Fair Isle as stepping stones? site axe from One thing is certain that the colonizing Shetland, must have been pulsating that means it ARC 65541. There is no presumably consisted of waves of influx indication securing a demographic success, or the either of wear Shetland population had either connec- or of hafting. Courtesy Ler- tions back to the mother population or wick Museum. was part of an exchange system as we have seen above with Sarqqaq and in Oceania. This brings us in search for one or more mediators or special prestige items. One would expect that the beau- tiful, point butted felsite axes, totally polished and some of them never used or hafted could be items in an exchange chain and thus a mediator (Mahler 2011b: 61), fig. 5. On Shetland these long special axes are often found in wetlands, and seen from an archaeological point of view, they are likely to be ritual deposits though we know precious little about the circumstances. But there are none of these axes outside Shetland – other than in museum collections (Com- pare Ballin this volume). The same goes es, and finds from Jarlshof and the foot for the Shetland Knives, which are found marks from Clickhimin suggest Westerly packed together in numbers up to 19 connections, maybe Irish (Turner 1998: and deposited in wetlands or in special 65). Whether these connections have places (Brandly 2000). As far as is known older roots is not yet known, and the we have no objects with affinity to Shet- proof of a supposed network during the land found in known Neolithic context Neolithic period is still to be found. Dur- on Orkney, Scotland or the Hebrides ing the Late Bronze Age renewed con- (Ballin 2011: 32 f.). The Bronze Age on tacts are made, and Shetland is again Shetland is very poor on actually bronz- part for much larger cultural area.

Nordlige Verdener Shetlandsprojekt 140

Acknowledgements Ballin, T. B. 2011: The Post-Glacial Colo- Bjarne Grønnow has been of great help nization of Shetland – Integration or Iso- with discussions about the Greenlandic lation. Evidence from Lithic and Stone material and Poul Otto Nielsen, who is a Assembleges. Farming on the edge: Cul- lifelong researcher into the Neolithic, tural Landscapes of the North. Some has helped in many other ways. Thanks features of the Neolithic of Shetland. to Martin Appelt for important referenc- Short papers from the network meeting es – all of them the National Museum of in Lerwick, Shetland September 7th-10th Denmark. Hans Christian Gulløv has pro- 2010. Northern Worlds. Mahler, D. L. & vided important support since the begin- Andersen, C. (eds.), København: 32-43. ning of Northern World projects at The The National Museum of Denmark. National Museum of Denmark. And a special thanks to my wife Susan Dall Bradley, R. 2000: An Archaeology of Mahler, who is a very patient person. Natural Places. London

Edwards, K.J., Schofield, J.E., Whitting- Bibliography ton, G. and Melton, N.D., 2009: Palynology ”On the Edge” and the Archaeological Anderson, A. 2001: Mobility models of Vindication of a Mesolithic Presence? The Lapita mirgration, in Clark, G. R., Ander- case of Shetland. From Bann Flakes to son, A. J. and Vunidilo, T. (eds.), The Ar- Bushmills, papers in honour of Professor chaeology of Lapita Dispersal in Ocean- Peter Woodman, Finlay, N., McCartan, S., ia. Papers from the Fourth Lapita Milner, N. and Wickham-Jones, C. (eds.), Conference, June 2000, Canberra, Aus- Prehistoric Society Research No 1, Ox- tralia. Terra Australis 17 Canberra: 15-23. ford, Oxbow: 113-123.

Anderson, D. G. & Gillam, J. C. 2000: Edwards, K.J. & Whittington, G. 1998: Paleoindian Colonization of the Ameri- Landscape and environment in prehistoric cas: Implications from an Examination West Mainland, Shetland. Landscape of Physiography, Demography and Arti- History. Vol. 20: 5-17. fact Distributions, American Antiquity 65: 43-66. Firth, R. 1959; Social Change in Tikopia, Re-study of a Polynesian Community after Andreasen, C. 2004: Independence 1-kul a Generation, London. London School of turen, Grønlands forhistorie, Gulløv, H. Economics. C. (ed.), Gyldendal, København: 37 - 64. Friedman, J. 1981: Notes on structure and history in Oceania, Folk 23: 275-295. 141

Grønnow, B. 2004: Saqqaqkulturen, Lynnerup, N. 1998: The Greenland Norse. Grønlands forhistorie, Gulløv, H. C. (ed.), A biological-anthropological study. Med- Gyldendal, København: 66 - 108. delelser om Grønland: Man and Society 24, København: 1-149.The Comission for Grønnow, B. & Sørensen, M. 2006: Scientific Research in Greenland. Paleo-Eskimo Migrations into Greenland: The Canadian Connection, Dynamics of Lynnerup, N. 2011: When population de- Northern Societies, Proceedings of the cline. End period demographic and eco- SILA/NABO Conference on Arctic and nomics of the Greenland Norse, Über die North Atlantic Archeology, Arneborg, J. Grenzen und zwischen den Disziplinen. & Grønnow, B., (eds.), København: 59-74. Fächerübergreifende Zusammenarbeit im Forschungsfeld historischer Mensch- Grønnow, B. IP.: Independence I and Umwelt-Beziehungen, Meier, T. und Tilles Saqqaq: the first Greenlanders, Hand- sen, P. (Hrsg.), Archaeolingua Alapituany book of Arctic Arhaeology, Friesen, M. & Budapest: 319-329. Mason, O. (eds.),Oxford University Press, Oxford: 1-23. Mahler, D. L. 2007: Sæteren ved Argis- brekka. Økonomiske forandringer på Irwin, G. 2006: Voyaging and Settlement, Færøerne i vikingetid og tidlig middelal- Vaka Moana, Voyages of the Ancestors, the der. Economic development during the Vi- Discovery And Settlement of the Pacific, king Age and Early Middle Ages on the Howe, K. R., (ed.), Hong Kong: 56-99. Faeroe Islands. Annales Societatis Sci- entiarum Færoensis Supplementum, Kirch, P. V. 1988: Long-distance exchange Faroe University Press, Tórshavn. and island colonization. The Lapita case. Norwegian Archaeological Review 21, Mahler, D. L. 2011a: Shetland: The Bor- No. 2: 103-117. der of Farming 4000-3000 BC. Features of the Neolithic Period on Shetland. Kirch, P. V. 2000: On the Road of the Winds: Farming on the edge: Cultural Land- An Archaeological History of the Pacific scapes of the North. Some features of Island Before European Contact. Berkeley the Neolithic of Shetland. Short papers University of California Press. from the network meeting in Lerwick, Shetland September 7th-10th 2010. North- Kirch, P. V., 2007: Concluding Remarks in ern Worlds. Mahler, D. L. & Andersen, C. Kirch, P. V. and Rallu, J-L. (eds.), The (eds.), København: 6-18. Growth and Collapse of Pacific Island So- cieties. Honolulu: 326–337.

Nordlige Verdener Shetlandsprojekt 142

Mahler, D. L. 2011b: Shetlandsøerne: Gillam, American Antiquity, 66: 526-529. Landbrug på grænsen 4000-3000 f.v.t. Neich, R. 2006: Voyaging after the Ex- Nogle træk af yngre stenalder på Shet- ploration Period, Vaka Moana, Voyages landsøerne. Ændringer og udfordringer, of the Ancestors. The Discovery and Set- rapport fra Workshop 1 på National- tlement of the pacific, Howe, K. R. (ed.), museet, 29. september 2010, Nordlige Hong Kong: 198-245. David Baeman, Verdner, Gulløv, H. C., Paulsen, C. og Auckland Museum, Auckland. Rønne B., (red.), København: 59-68. Robert-Lamblin, J. 2006: Demographic Malinowski, B. 1922: Argonauts of the Fluctuations and Settlement Patterns of Western Pacific. Percy Lund, Humphries the East Greenlandic Population – as & Co, London Gathered from Early Administrative and Etnographic Sources. Dynamics of North- Melton, N.D. 2009: Shells, seals and ce- ern Societies, Proceedings of the SILA/ ramics: an evaluation of a midden at NABO Conference on Arctic and North At- West Voe, Sumburgh, Shetland, 2004-5. lantic Archeology, Arneborg, J. & Grøn- Mesolithic Horizons, Papers presented at now, B., (eds.), København: 235-246. the Seventh International Conference on the Mesolithic in Europe, Belfast 2005. Rowley-Conwy, P. IP: Westward Ho! The MaCartan, S., Schulting, R., Warren, G. Spread of agriculture from Central Europe to and Woodman, P. (eds.), Vol. 1, Oxford, the Atlantic, Current Anthropology: 1-29. Oxbow: 184-189. Sheridan, A. 2010: The Neolithization of Melton, D.N. & Hicholson, E.A. 2007: A Britain and Ireland: The “Big Piture”, Land Late Mesolithic – Early Neolithic midden scapes in Transition, Finnlayson, B. & at West Voe, Shetland. Shell Middens in Warren, G. (eds.), Oxford, Oxbow: 89-105. Atlantic Europe. Milner, N., Craig, O. E. and Baily, G. N. (eds.), Oxford, Oxbow books. Sørensen, M. & Pedersen, K. B. 2005: Killiaq kilden på Nuussuaq: på sporet af Moore, J. H. 2001: Evaluating Five Mod- Saqqaqkulturens redskabssten. Tidsskrift- els of Human Colonization. American et Grønland nr. 2-3, 2005: 105-118. Anthropologist 103 (2): 395-408. Turner, V. 1998: Ancient Shetland. Histor- Moore J. H. & Moseley M. E. 2001: How ic Scotland. Shetland Amenity Trust. D. many Frogs does is take to Leap around J. Breeze (ed.), Batsford Ltd. London the Americas? Comments on Anderson and 143

The heel shaped cain on the island of Vementry.

The prehisto ric house at Punds Water.

Photo Ditlev L. Mahler.

Nordlige Verdener Shetlandsprojekt 144

Contributors

Torben Ballin Will Megarry PhD PhD Candidate Lithic Research University College Dublin [email protected] [email protected]

Gabriel Cooney Jenny Murray Professor Curator of Collections, University College Dublin Shetland Museum and Archives [email protected] [email protected]

Hans Christian Gulløv Emmett O’Keeffe Senior researcher, DPhil, PhD Candidate Etnografical Collection, University College Dublin The National Museum of Denmark [email protected] [email protected] Preben Rønne Flemming Kaul Associate professor PhD Senior researcher, DPhil, Vitenskapsmuseet, Danish Prehistoric Collection, University of Trondheim, The National Museum of Denmark [email protected] [email protected] Alison Sheridan Ditlev L. Mahler Dr, FSA Scot FSA AIFA Project senior researcher, PhD Principal Curator Danish Prehistoric Collection, Department of Scottish History The National Museum of Denmark, and Archaeology Coordinator National Museums Scotland [email protected] [email protected]

Stephen Mandal Graeme Warren PhD Lecturer, PhD Managing Director, CRDS University College Dublin [email protected] [email protected] The Border of Farming and the Cultural Markers

Backcover: The audience at the day of papers: To the left of the boat from the front: Esther Renwick, Preben Rønne, Ditlev L. Mahler, Gabriel Cooney, Chris Dyer. To the right of the boat from the front: Jenny Murray, Lauren Daughton, Deborah Lamb, Simon Clarke, Ian Tait, Alison Sheridan, Will Megarry, John Hunter. The group to the right from behind the left: Val Turner, Flemming Kaul; and in front from the left Carol Christiansen and Laurie Goodlad.

The National Museum of Denmark Frederiksholms Kanal 12 DK-1220 Copenhagen K http://nordligeverdener.natmus.dk

The Border of Farming and the Cultural Markers

Short papers from the network meeting in Lerwick, Shetland September 5th – 9th 2011