by Chris Fowler
Noble, G., Grieg, M., and Millican, K. 2012. Excavations at a multi-period site at Greenbogs, Aberdeenshire, Scotland and the four-post timber architecture tradition of late Neolithic Britain and Ireland. Proceedings of the Prehistoric Society 78, 135–172.
Ó Nualláin, S. 1972. A Neolithic house near Ballyglass at Ballycastle, Co Mayo. Journal of the Royal Society of Antiquaries of Ireland 106, 92–117.
Ó Ríordáin, S.P. 1954. Lough Gur excavations: Neolithic and Bronze Age houses on Knockadoon. Proceedings of the Royal Irish Academy 56C, 297–459.
Oxford Archaeological Unit. 2000. White Horse Stone: a Neolithic longhouse. Current Archaeology 168, 450–453.
Parker Pearson, M. and Richards, C. (eds) 1994a. Architecture and order: studies in social space. London: Routledge.
Parker Pearson, M. and Richards, C. 1994b. Ordering the world: perceptions of architecture, space and time. In M. Parker Pearson and C. Richards (eds), Architecture and order: studies in social space, 1–37, London: Routledge.
Parker Pearson, M., Pollard, J., Richards, C., Thomas, J., Tilley, C., and Welham, K. 2006. The Stonehenge Riverside Project. Summary interim report on the 2006 season. Available at www.shef.ac.uk/archaeology/research/stonehenge
Pollard, J. 1999. ‘These places had their moments’: thoughts on settlement practices in the British Neolithic. In J. Bruck and M. Goodman (eds), Making places in the Prehistoric world: themes in settlement, 76–93, London: University College London Press.
Richards, C. 1990. The late Neolithic house in Orkney. In R. Samson (ed), The social archaeology of houses, 111–124, Edinburgh: Edinburgh University Press.
Richards, C. 2005. Dwelling amongst the monuments. The Neolithic village of Barnhouse, Maeshowe Passage Grave and surrounding monuments at Stenness, Orkney. Cambridge: McDonald Institute Monograph.
Ritchie, A. 1983. Excavations of a Neolithic farmstead at Knap of Howar, Papa Westray, Orkney. Proceedings of the Society of Antiquaries of Scotland 113, 40–121.
Samson, R. (ed.) 1990. The social archaeology of houses. Edinburgh: Edinburgh University Press.
Saville, A. 1990. Hazelton North: the excavation of a Neolithic long cairn of the Cotswold–Severn group. London: English Heritage.
Simpson, D.D.A. 1996. Excavation of a kerbed funerary monument at Stoneyfield, Raigmore, Inverness, Highland, 1972–3. Proceedings of the Society of Antiquaries of Scotland 126, 53–86.
Simpson, D.D.A. and Coles, J.M. 1990. Excavations at Grandtully, Perthshire. Proceedings of the Society of Antiquaries of Scotland 120, 33–44.
Smyth, J. 2006. The role of the house in early Neolithic Ireland. European Journal of Archaeology 9.2/3, 229–257.
Smyth, J. 2010. The house and group identity in the Irish Neolithic. Proceedings of the Royal Irish Academy 111C, 1–31.
Thomas, J. 1996. Neolithic houses in mainland Britain and Ireland—a sceptical view. In T. Darvill and J. Thomas (eds), Neolithic houses in northwest Europe and beyond, 1–12, Oxford: Oxbow.
Thomas, J. 1999. Understanding the Neolithic. London: Routledge.
Thomas, J. 2006. On the origins and development of cursus monuments in Britain. Proceedings of the Prehistoric Society 72, 229–241.
Thomas, J. 2010. The return of the Rinyo-Clacton folk? The cultural significance of the Grooved Ware complex in Later Neolithic Britain. Cambridge Archaeological Journal 20, 1–15.
Toolis, R. 2011. Neolithic domesticity and other prehistoric anomalies: excavations at Laigh Newton, East Ayrshire. Scottish Internet Archaeology Reports 49. Available at http://archaeologydataservice.ac.uk/archives/view/sair/contents.cfm?vol=49
Topping, P. 1996. Structure and ritual in the Neolithic house: some examples from Britain and Ireland. In T. Darvill and J. Thomas (eds), Neolithic houses in northwest Europe and beyond, 157–170, Oxford: Oxbow.
Whittle, A., Keith-Lucas, M., Milles, A., Noddle, B., Rees, S., and Romans, J.C.C. 1986. Scord of Brouster. An early agricultural settlement on Shetland. Oxford: Oxford University Committee for Archaeology.
Whittle, A. 1988. Problems in Neolithic archaeology. Cambridge: Cambridge University Press.
Whittle, A. 2003. The Archaeology of people: Dimensions of Neolithic life. London: Routledge.
CHAPTER 18
PLACES OF SETTLEMENT IN SOUTHERN SCANDINAVIA
MATS LARSSON
INTRODUCTION
DURING the past couple of decades our understanding of the Neolithic settlement structure and how people used their landscape in southern Scandinavia has fundamentally changed. The foundation for this marked change in our understanding of how Neolithic communities organized their settlements is to a large degree based on the development of contract archaeology. Large-scale excavations, stretching in several cases over years, have given us evidence for complex Neolithic settlements that was unheard of twenty-five years ago. There is now a well-defined house typology with clear chronological implications for this region. This chapter discusses Neolithic settlement, focusing on chronological and regional patterns in the design and use of buildings, the spatial arrangement of settlements, the landscape location of settlements, and activities at these places.
THE END OF LATE MESOLITHIC SETTLEMENTS
The large coastal sites inhabited during the Ertebölle period (c. 5400–4000 BC)1 were more or less abandoned as permanent settlements after having been in use for about 1,000 years. For instance, the Björnsholm shell midden in northern Jutland was in use from c. 5050–4050 BC (Andersen 1993, 59–61). The only substantial features found in the midden were hearths: there were no traces of huts, pits, or postholes. Yet late Mesolithic houses and huts have recently been excavated at Tågerup in western Scania (Karsten and Knarrström 2001, 83ff). The chronology of these houses is complicated as there are few Mesolithic radiocarbon dates (Karsten and Knarrström 2001, 147), meaning that dating has relied on a study of the transverse arrowheads (Vang-Petersen 1984). The houses were inhabited during the early and middle phases of the Ertebölle culture, whilst the presence of Ertebölle ceramics from House III suggests a late Ertebölle date (c. 4600–4000 BC) (Karsten and Knarrström 2001, 143; cf. Hallgren 2004, 136). These middens often show continued periodic use after c. 4000 BC. Inland settlements are also known from many parts of southern Scandinavia (cf. Fischer 2002, 369; Larsson 1982, 1983; Karsten 2001).
EARLY NEOLITHIC (FUNNEL BEAKER (TRB): C. 3950–3300 BC)
In the early Neolithic, people’s view of their environment—and accordingly their world view—changed, resulting in the building of earthen longbarrows and other types of monuments as well as new types of houses. It is also interesting to note that the same types of houses were built at the same time by people who were using different types of pottery (Oxie, Volling, Svenstorp and so forth; see Petersen and Müller, this volume). People during the late Mesolithic built dwellings in a diversity of ecological areas ranging from the fjords of eastern Jutland to the inland of Scania, whilst in the earliest part of the early Neolithic new areas inland, preferably away from the late Mesolithic sites, were occupied, preferably sandy soils close to water (cf. Madsen 1982; Larsson 1984). These early settlements sites have usually been seen as small, but this notion is under discussion as new evidence from, for example, Dösjebro in Scania indicates the existence of several potentially contemporaneous houses (cf. Larsson 1984; Kyhlberg et al. 1995; Andersson 2003, 75).
Many of the earliest TRB sites consist of different types of pits, sometimes in their hundreds, and little else (Larsson 1984, 1985). These pits were frequently recut and re-used, and contained large amounts of flint debris as well as unused implements like flake axes, flake scrapers, and in some cases even complete vessels (M. Larsson 1984). Several of these sites were situated on ridges or small hills in the undulating landscape. If we accept these places as not only ordinary settlement sites but as places of special virtue for the people of the day, we can see these places as evidence for an early ‘monumentalization’ of the landscape (Larsson 2007a).
Houses of diffe
rent types have also been excavated in both Denmark and southern Sweden, though most of these early excavations that mention houses or huts are hard to interpret and most of the Danish ones were also to a large degree discarded in an article concerning house construction some years ago (Eriksen and Madsen 1984). Classic examples of re-interpretation are Barkaer and Stengade in Denmark, which were both interpreted as longhouses but were later convincingly re-interpreted as longbarrows (Skaarup 1975; Madsen 1979; Liversage 1992). Probably the first time a proper longhouse was excavated was in 1986 at Mossby, in the most southern part of Scania (M. Larsson 1992). The Mossby house, now a ‘type site’, was c. 12 by 6m in size, with three large postholes in the centre to support the roof. The remains of the wall were made up of rather shallow postholes and the form of the house was elliptical. A layer consisting of sooty sand and occupational debris covered an area of about 70sq. m, largely within the walls and fitting the shape of a floor depression. There were few artefacts, but these included cord-decorated pottery and flint waste. Some 10m east of the house a 280sq. m area was excavated, rich in occupation debris, including pottery and flint. Beneath the debris pits, patches of soot and postholes were found, but it was not possible to delimit any houses in this area (M. Larsson 1992, 66; Larsson and Rzepecki, 2005). The pottery found in the layer and in the pits was identical to that found in the layer that covered the house. The five available radiocarbon dates are all in the range of 4100–3900 BC, putting them among the earliest in southern Scandinavia (M. Larsson 1992, 74). These dates have been doubted, especially the oldest ones, which derive from charred food residue (Persson 1999).
Of the 36 houses currently known from early and middle Neolithic southern Scandinavia 17 are Mossby-type houses, and these can be dated from the earliest Neolithic up until the early middle Neolithic (Svensson 2003, 123). The size of these buildings varies between 36 and 128sq. m. Houses of this type have been excavated at several places on Öland, in eastern central Sweden (Papmehl-Dufay 2010; Larsson 1994; Carlsson 2004; Ahlbeck and Artursson 1996; Hallgren et al. 1997; Hallgren 2008; Stenvall 2007). For instance, at the site Skogsmossen in the county of Västmanland a house of Mossby type was identified on the basis of a stone-free area devoid of finds within areas of occupation debris (Hallgren et al. 1997; Bradley 2005).
Large-scale rescue excavations have for many years been important in bringing forward new material. In connection with the extensive excavations performed in western Scania for a new railway at the end of the 1990s a large number of early Neolithic settlements were excavated. On several of these sites, houses have been found. In the vicinity of the village of Dagstorp outside the city of Lund, on slightly undulating farmland along a river valley, probably the largest number of Neolithic houses in southern Scandinavia has been excavated (Andersson 2003, 2004; Svensson 2003; Fig. 18.1). The settlements make up a sequence from the earliest Neolithic to the central part of the middle Neolithic. Different types have been distinguished: Mossby-type houses, and three others that will be discussed later; Dagstorp-type 1 houses, Limensgård houses, and Dagstorp-type 2 houses (Svensson 2003). The latter type of houses belongs to MN A and MN B, and these are basically derivatives of the Mossby house. A number of what have been interpreted as huts were also found. Most of the early Neolithic houses found at Dagstorp and elsewhere were situated close to or were covered by a layer of occupation debris, and pits were common features of these settlements (Svensson 2003, 75). The large scale excavations have demonstrated that settlements could be very large: up to several thousand square metres (Svensson 2003, 132). The Mossby house was very clean—a feature matched elsewhere—and we can infer that most activities took place outside the houses (M. Larsson 1992). It is of course also possible that some activities took place inside the house but debris was removed and deposited elsewhere.
FIG. 18.1. Early Neolithic houses at Dagstorp, western Scania (Andersson 2004).
Settlement evidence from Denmark is rather different. In most cases the excavated areas around the structures are too small to draw comparisons with the Swedish evidence (cf. Buus-Eriksens 1992), but there are exceptions, like the Mossby-type house excavated at Skraeppekaergård in northern Zealand (Kaul 1997). Areas of occupation debris (including pottery and flint) were situated 4–8m from the house and covered areas of 25 by 25m and 20 by 20m.
The picture that we today have of early Neolithic settlements is based on excavations in a rather small number of well-researched areas like southern Sweden, parts of central Sweden, the island of Bornholm, and parts of Denmark. We can discern several types of houses, such as, for example, Mossby houses, D-shaped huts and horseshoe-shaped features. There seem to be some regional—and local—variation but the evidence for this is inconclusive. The size of the sites varies but this may be a factor of the scale of different excavations.
THE MIDDLE NEOLITHIC
The evidence for houses of any kind from the middle Neolithic (c. 3300–2350 BC) is scant in comparison with the preceding period (Larsson 1995; Larsson and Olsson 1997; Svensson 2003, 115). The middle Neolithic is usually divided into MN A (c. 3300–2700 BC: TRB and Pitted Ware culture (PWC)) and MN B (c. 2700–2350 BC: Battle Axe culture (BAC) and Single Grave culture (SGC)). The literature regarding the different culture groups and their relationship is today vast and almost impossible to review (cf. Edenmo et al. 1997; Malmer 2002; Larsson 2006; Papmehl-Dufay 2006; and see Thorpe, Rowley-Conwy and Legge, and Petersen and Müller, all this volume) but, in brief, the TRB, BAC and SGC have been linked with farming and animal husbandry, whilst the PWC has been seen as subsisting by hunting and gathering.
Middle Neolithic A
The single-aisled houses of the Mossby type obviously have a long time span, from EN up until the beginning of MN A (Svensson 2003, 116), but from this point on very few Mossby-type houses have been found. During the very extensive excavations in the Malmö area in the 1990s and 2000, no single house datable to this period has been found. Indeed, in most areas of southern Scandinavia there are no houses from the earlier part of the MN A. The first houses in the area belong to MN B, leaving a long hiatus. During the late MN A the situation changed, with a few excavated houses from southern Sweden. From the extensive excavations at Dagstorp mentioned earlier we have a couple of MN A houses associated with TRB artefacts. The Scanian houses have no wall trenches and the houses are orientated east–west or north-west/south-east, c. 13.5 to 19m long by 5 to 7m wide, and with between three and six internal roof supports (Svensson 2003, 116ff).
During the large excavations at Limensgård and Grödbygård on Bornholm several houses of different types and age have been documented (Nielsen and Nielsen 1985; Kempfner-Jörgensen and Watt 1985; Nielsen 1999). The sites are dated with radiocarbon to between c. 2900–2700 BC. Mossby, Dagstorp and Limensgård type houses are present. The difference between them is sometimes subtle. The settlements have a long duration and have been extensively used from the early Neolithic up until the late Neolithic. At Grödbygård two late MN A houses were excavated. House A measured c. 13 by 7m. The rectangular outline of the house was shown by a partially preserved wall trench, and three roof supporting posts ran along the axis. House B was badly preserved and lacked any trace of a wall trench, but had the same three roof supporting posts and was no more than 10–11m long (Kempfner-Jörgensen and Watt 1985, 87ff). At Limensgård one late MN A house (Y) was excavated (Nielsen and Nielsen 1985, 102f). This house is quite similar to the ones from Grödbygård. The full length of this house cannot be determined; however its preserved portion is about 18 by 6.2m. There are five central postholes and a narrow wall trench was preserved in the northern part.
To summarize, the number of houses from early MN A is small and the houses of the Mossby type seem to prevail. At the end of the period the number of houses is larger and a new type of house defined by a wall trench is introduced. How then can we interpret these changes? During late MN A there is a lot of evidence for a rapidly changing TRB society in which the use of, for example,
passage graves stopped. We can also see that in some well-researched areas in Scania it is possible to see marked changes in settlement structure as early as MN III (the middle part of MN A). In many areas we have been able to trace continuous TRB settlement from at least the late EN up until MN III, followed by a rapid change around 2700 BC when the old settlement regions were partly abandoned and new regions came into use (M. Larsson 1992; Andersson 2004).
Pitted Ware culture huts and houses date from the MN A into the MN B. The youngest radiocarbon dates from PWC sites in eastern central Sweden are 2880–2490 BC from Åby and 2580–2450 BC from Bollbacken. These stretch well into MN B (Artursson 2006, 50). At Åby a hut c. 5.5 by 2m was excavated in 1992 (M. Larsson 1999, 2003, 2006). Beneath a layer with red-brown sand several features, such as pits, postholes, and what has been interpreted as pit house, were found. A number of postholes were found, and in the close vicinity of the structure at Bollbacken four post-built huts, 2 by 6m in size, were documented (Artursson 2006, 50ff). Each hut was associated with activity areas, cooking pits, postholes and hearths (Artursson 2006, 50f). In House 3, as well as in features associated with it, cremated human bone and animal bones were found. The excavator interprets the construction as a ritual building with an associated area around it that was used for ritual activities like funerary rites (Artursson 2006, 55f). What have been interpreted as ritual areas, pits, and structures are not that uncommon on Pitted Ware sites (Larsson 2007b). Further to the north at Fräkenrönningen a settlement, radiocarbon dated to 3090–2390 BC, resembled the site at Bollbacken in that six of the huts were situated in a semi circle around an area more or less empty of finds (Björck 1998; Fig. 18.2). The site covers an area of about 50 by 30m. In this area seven, possibly eight, hut structures were excavated. They were all visible as semicircular stone foundations and soil colorations (Björck 1998; 21). The hut circumferences varied between 5.5 and 14.7m.
