The Neolithisation of Scandinavia How Did It Happen?

Niels V. Skak-Nielsen The neolithisation of Scandinavia How did it happen?

Abstract Since the early 1970s there has been a near-consensus among archaeologists that agriculture was introduced to southern Scandinavia around 4000 cal BC without any immigration, through a voluntary decision by the indigenous hunter-ﬁsher population. It has been sur- mised that the necessary technology was adopted through contact with Neolithic cultures on the Continent. In the present paper it is pointed out that these views, as well as later models of an economic shift without any immigration, are untenable. In the absence of any immigration, there would have been neither opportunity nor motive for the shift to a new economic culture. The European background to the shift is reviewed, taking into accout new results in this ﬁeld of research. It is shown on the basis of published archaeological research that there has been immigration into southern Scandinavia. The expansion of Neolithic cultures took c. two centuries from Holstein into the peninsula of Jutland and the Danish Isles, onward to Scania, and then to Bornholm, the Lake Mälaren area, Gotland and the Oslo Fiord. There is a strong likelihood that a climate change at the time, which caused the Danish straits and part of the Baltic Sea to freeze over during the winters, made this rapid spread of a new culture over such a large area possible. It is shown that this spread entailed immigration on a considerable scale compared to the size of the area’s Mesolithic population. Key words: Scandinavia, Neolithic, immigration, climate change

Siden begyndelsen af 1970’erne har der blandt arkæologer bestået næsten konsensus om, at landbrug omkring 4000 f.Kr. blev indført i den sydlige del af Skandinavien uden indvandring gennem en frivillig beslutning af den derværende jæger- og ﬁskerbefolkning. Man antog, at den nødvendige teknologi var blevet tilegnet gennem kontakter med neolitiske kulturer på kontinentet. Det påpeges, at disse synspunkter – såvel som senere fremførte forklaringer på et erhvervsskifte uden indvandring – ikke kan være holdbare. Uden indvandring har der hverken været mu- lighed for eller tilskyndelse til overgang til den nye erhvervskultur. Den europæiske baggrund for erhvervsskiftet gennemgås under inddragelse af resultaterne af ny forskning på området. At indvandring i det sydlige Skandinavien faktisk må have fundet sted påvises på grundlag af eksisterende arkæologisk forskning. Spredningen af de neolitiske kulturer skete indenfor et par hundrede år fra Holstein over den jyske halvø og de danske øer til Skåne og derfra til Bornholm, Mälardistriktet, Gotland og til Oslofjorden. Det er overvejende sandsynligt, at en samtidig ændring i klimaet og deraf følgende tilfrysning om vinteren af farvandene mellem de danske øer og Sverige og i Østersøen har muliggjort den hurtige spredning af den nye kultur over et meget stort om-

89 råde. Det påvises, at der må have været tale om en betydelig indvandring målt i forhold til størrelsen af den hidtidige mesolitiske befolkning. Nøgleord: Skandinavien, neolitikum. immigration, klimaændring

Introduction: The Danish-Swedish synthesis Until about 1970 the dominant view among that interpreted the neolithisation of Europe Danish and Swedish archaeologists was that as the colonisation by a farming population the Neolithic came to Denmark and southern of an area inhabited by Mesolithic fishers Sweden through an immigration of farm- and hunters. ers from the south. In many other cases, Among Danish archaeologists, the reason for migration had been used incorrectly as an the new view of the neolithisation was that explanation of changes in the archaeologi- radiocarbon dates in the 1970s appeared to cal record, and these explanations had been show that the final phase of the Mesolithic misused for political purposes (e.g. Nazi ideas Ertebølle culture (EBK) had been replaced of Aryan migration). Therefore a change almost without any chronological overlap by occurred in most archaeologists’ attitude a farming culture that was called the Funnel to the question of neolithisation. This was Beaker culture (TRB) due to its characteristic partly due to the New Archaeology. Migra- pottery. This led to the conclusion that there tion was generally discounted as an explana- had been no immigration. In such a case, the tory model and evolutionist or diffusionist new culture would have had to co-exist with ideas were preferred. According to these, EBK for a considerable time before the Me- the observed changes would either have solithic population had been assimilated or been caused by evolution within a society, by exterminated. There appeared to have been ecological parameters or by foreign cultural no period of such “culture dualism”. Thus influences. A belief in migration was largely the TRB population was seen as immediate seen as politically incorrect. There may also descendants of the Ertebølle people. have been an element of nationalism in this However, one may just as well argue the op- changed viewpoint, and the insurrection of posite. If the shift involves not only culture as 1968 against “old professors” would also seen e.g. in mortuary customs, but also basic have played a part. Some scholars may have technologies (agriculture, stock breeding, been motivated by a wish to paint an idyllic flintaxe grinding and new pottery), then it picture of the past. must take quite some time to develop locally, The New Archaeology originated among while it would be easier to imagine an abrupt Anglo-American archaeologists who shared shift if an immigrant farming population this aversion toward explanations based on replaced the previous Mesolithic one. Swed- migration. Many British archaeologists tend ish archaeologists, who agreed with their to deny any prehistoric immigration into the Danish colleagues in denying immigration, UK after the original one that took place took the latter view. The shift appeared to while the British Isles were land-locked for have been more gradual in Scania than in a time after the latest Ice Age. History docu- Denmark, and this was taken also to indicate ments at least four migrations to the Isles an absence of immigration. in later millennia. Against the background of the current movement toward European unity, it may have been important that the The Continental background UK, like Scandinavia, is on the periphery of It is commonly accepted that agriculture our subcontinent. The wish to distance one- reached Europe in Greece from Anatolia c. self from the recent colonial past beyond 7000 cal BC. About 1500 years later it had Europe has clearly also contributed to the reached present-day Hungary across the resistance against any explanatory model Balkans (Price 2000). Here the Linear Pot-

90 tery Culture (German Linearbandkeramik, economy suited to the less fertile, mainly LBK) appeared, and during two centuries forested moraine areas of northern Germany. spread westward, through Germany across The eventual northward expansion of TRB the great plains of loess soil to parts of the from Holstein took place with a speed and Netherlands and Belgium, and eastward, to across distances that can be compared to the Ukraine. the expansion of LBK across Central Europe Another branch of the Neolithic movement and of Neolithic cultures across the western spread westward from Greece across the reaches of the Mediterranean. Mediterranean. About 5500 cal BC it had But there are also examples of a slower reached the east coast of Italy. Then it began spread of Neolithic culture traits. Before the to move fast. It reached Portugal in about appearance of TRB, a slow diffusion of LBK two centuries, following the coasts of Italy technology can be seen in northern Germany and southern France and around the Iberian and Holstein in the first half of the 5th mil- peninsula. A branch reached up from the lennium cal BC (Hartz et al. 2002). Bones of coast through the Rhone river valley. From domesticated cattle dated to 4800–4600 cal Portugal it apparently continued along the BC have been found in Mesolithic culture Atlantic coast and formed the Michelsberg layers at Rosenhof in Holstein. There are culture in Northern France, Belgium and also indications of cereal agriculture north of the Rhineland during the second half of the LBK’s limit at the time in northern Germany. 5th millennium cal BC. Part of this culture Agriculture and stock breeding nevertheless was influenced by the LBK that had reached appears to have been of little economic im- western Europe long before from the east, portance at this stage. The spread of these as described above. The resulting mixed cul- Neolithic technologies would probably not ture rapidly spread back eastward through have been accompanied or caused by mi- Central Europe (cf. Klassen 2004:273). These gration. Only with the spread of TRB across two main Neolithic groups appear to have southern Scandinavia about 4000 cal BC did pushed each other to and fro, probably the entire Neolithic culture complex expand through migration. rapidly, which, as argued below, must have The majority opinion among scholars seems had to do with migration. to be that both the rapid expansion of LBK Where did the migrants come from during across Europe c. 5500 cal BC and the Medi- the rapid expansion of the LBK and later terranean expansion would only have been by the TRB? The most likely answer seems possible through migration. Some of the to be that the population expansion that areas reached by these expansions appear followed upon neolithisation first claimed never to have been populated before the nearby marginal soils. More labour intensive arrival of Neolithic populations. Many, how- agriculture could also put food on the table ever, also believe that the spread of LBK and for a time as populations grew. Eventually, of Neolithic cultures in i.a. parts of France however, emigration would have begun to and Germany were a result of the Meso- look more and more appealing. lithic populations’ acceptance of the new culture, perhaps after a period of reciprocal acculturation. Arguments against the immigration hy- After 5200 cal BC the geographic limit of neo- pothesis lithisation in northern Germany was largely Dating problems stable for more than a millennium. Mesolithic The original argument against the immigra- population density was relatively high here tion hypothesis – the new datings – is, as dis- in the coastal areas – as was the case in south cussed above, untenable. Furthermore, many Scandinavia. The Ertebølle culture was based of the datings from the transitional phase on a solid fisher-hunter-gatherer economy. turned out to have large margins of error In the later part of this period the Funnel when tree-ring calibration became available Beaker Culture (TRB) developed slash-and- in the 1980s. The rather flat calibration curve burn agriculture, an adaptation of the LBK at the transition to the Neolithic in southern

91 Scandinavia gives particularly large margins farmers had already lived there for more of error (see below). Other sources of error than 1000 years. have also been identified. This idea was based on finds from Denmark (primarily Lolland and Zealand) and Scania Where did TRB come from? of “shaft-hole axes or adzes”, tools made of Another argument against the immigration amphibolite, a hard but not brittle mineral. hypothesis was that one could not point to Their function is a matter of debate. They any particular area south of Jutland from were used in LBK for about 1200 years and which the farmers would have come. In thousands have been found in the area of Jutland, as on Zealand and in Scania, there this culture. The shape and use wear of these are two Neolithic groups with differences tools, as well as details of their design, indi- in their material and ideational culture. The cate that they were not in fact used as axes Oxie group is most common on the Danish but as hoes. They would have worked well Isles and in Scania. The Volling group was to make furrows in the hard loess soil for initially dominant in Jutland. According to the seed corn and to remove weeds. A few new research (Klassen 2004:323, 339), the grisly finds show that they were also used Oxie group had links to areas in central Ger- as weapons. many. The roots of the Volling group appear to have been in the Michelsberg culture of The idea was that the trade contacts indi- the westernmost parts of Europe that had cated by the shoe-last axes could have built not been influenced by LBK. The problem of up local knowledge that enabled the neo- where these cultural influencies came from lithisation. The technology and culture was remains, regardless of whether migration available to neighbouring Mesolithic popula- or diffusion is seen as the impetus for the tions during an often very long “availability economic shift. phase” (Zvelebil & Rowley Conwy 1986). They would thus have been able to make the shift Continuous flint technology whenever they wanted to. As early as 1982, A third argument against the immigration Fischer suggested such an explanation. Along hypothesis was that there were not any great with many other scholars, he supposes that changes in flint technology apart from the the possession of shoe-last axes conferred appearance of ground axes. The agricultural prestige in south Scandinavian Mesolithic populations of Europe, however, lived for a societies, and that this was the reason that very long time alongside Mesolithic cultures. they were imported. The question, however, They may have acquired their neighbours’ is if even intensive trade contacts allow the superior flint technology, which was similar transferral of technology and changed cul- to that of the Scandinavian Mesolithic. tural traits. If the axes came to Denmark and Scania in the Late Mesolithic, they may have been traded from settlement to settlement How do opponents of the immigration hy- across small distances. pothesis explain the shift to TRB? Some Swedish archaeologists have had dif- The arguments against immigration are thus ficulties in applying this model to the spread untenable. How, then, do the opponents of of neolithisation northward from Scania. this hypothesis explain the economic shift? It is hard to imagine intensive contacts be- A requisite for the shift must have been that tween Mesolithic groups on the shores of the hunters and fishers of southern Scandi- Lake Mälaren and the Oslo Fiord and farm- navia learned how to cultivate the soil and ers south of the Baltic (cf. Kihlstedt et al. breed livestock, grind flintaxes and make the 1997:123). These scholars did not, however, new pottery. It was supposed that there had contest the applicability of the availability been a far-reaching trade network across the model to Denmark and Scania. To explain Baltic connecting Late Mesolithic southern the expansion to the Lake Mälaren area it Scandinavia with farming populations on was supposed that there had been a long- the Continent. Around 4000 cal BC these distance social network between this area

92 and Scania through which the new culture against the existence of seaworthy vessels propagated. already in the Mesolithic that could have Hallgren (1996) suggested that these two been used for long seaborne journeys. If areas formed a Late Mesolithic social unit such journeys had been possible with other whose component groups exchanged spouses types of vessels then there would have been across distances of 400–500 km, thus allow- no reason to improve the construction of ing the new technology to spread rapidly to the dugouts. If there were other, more sea- Lake Mälaren. Hallgren’s model was adopted worthy boat types in Mesolithic southern by e.g. Jensen (2001) and Fischer (2002) as a Scandinavia, then one would also expect possible subsidiary explanation for the spread finds of such boats from the following mil- of TRB to Denmark. There is no evidence of lennia – but there are none. Nor are such such spouse exchange networks in northern seaworthy craft known from other countries Europe. Price et al. (2001), however, argued in northern Europe at the time. for such exchange between farmers in the Rhineland and Mesolithic groups in adjoining Scandinavian archaeologists (cf. Hesjedal et mountainous areas. If there had been close al. 1996) have suggested that there may have contacts between Scania and Lake Mälaren been boats made of animal skins stretched in the Late Mesolithic, it is inexplicable that across a light-weight wooden frame, in such a useful technology as the pointed-base southern Scandinavia. Such craft are depicted pots and pottery lamps of EBK never spread in rock carvings prior to 4000 cal BC in the north of Scania (cf. Persson 1999:134). Nor circumpolar area. Some of them are similar did other elements of EBK spread north of to the umiak of the Inuit. They could carry Scania. several people but were only useful in coastal The availability model, as said before, re- archipelagos. No remains of such boats are quires intensive contacts between Mesolithic known from southern Scandinavia, and Neo- southern Scandinavia and north German lithic rock carvings south of the line Näm- farmers. There was, however, most likely forsen-Trondheim do not feature any boats a language barrier. The farmers south of (Lindkvist 1994:18). Boats are, however, very the Baltic would probably have spoken an- common in Bronze Age rock art. They may other language than the Ertebølle people (cf. represent a development of boat types after Renfrew 1987; Bellwood 2004). Be that as it 3000 cal BC leading towards the plank-built may, transportation would have posed even boats of the Iron Age. The earliest find of worse problems. Land transportation was such a boat is the Early Iron Age boat from difficult. All travel was by foot and entailed Hjortspring in Denmark (c. 300 cal BC). considerable risks. Times were anything but Our knowledge of Ertebølle material culture peaceful. Seaborne transportation was the in various parts of southern Scandinavia tells best way to move people and goods over us quite a deal about transportation and long distances. communication at that time. There is hardly Many Mesolithic log dugouts have been any difference between Zealand and Scania, found. The longest one, measuring more which means that the strait of Öresund was than 10 metres, dates from the late 5th mil- no barrier. The strait of Storebælt, however, lennium BC and was found at Tybrind Vig in was apparently a barrier against frequent Denmark (Andersen 1988). There are also contact as shown by the differences between many Early Neolithic dugouts from the 4th material culture on Zealand and in Jutland. millennium, but not from the later stages There is no sign of contact between Born- of the Neolithic in the 3rd and early 2nd holm and the Continent, where the distance millennium. is 90–100 km. Any seaborne contact in the The Neolithic dugouts have new technologi- Late Mesolithic across the 40 km between cal traits. Details indicate that planks were Bornholm and Scania was very limited (see sometimes added to raise the gunwale, which below). would make the vessel more seaworthy (Rieck There is thus no indication that the Mesolithic & Crumlin-Pedersen 1988). This fact argues population of southern Scandinavia could

93 have switched economies by themselves. They In the following we shall review further main would not, in the absence of close contact arguments that an immigration did in fact with farmers, have been able to acquire their take place. technology or cultural traits.

Another question is why these fishers and Further main arguments for the immigra- hunters would have chosen to switch econo- tion hypothesis mies. Many archaeologists believe that the Problems with diffusion Mesolithic economy was not very labour An important argument for the immigration intensive. This view is based on, among other hypothesis is that a diffusion of technology things, anthropological studies of modern from one area to another would have been societies with a similar way of life. Thus, an extremely slow process. This has been the the economic shift was not inescapable but case also in historic times. In the early 16th must be explained. Some have suggested century AD, King Christian II of Denmark ecological and population-based factors be- wanted his people to grow greens like the hind the shift: overpopulation, elm decline Dutch did, for the provision of Copenhagen. and climate deterioration, including a local For this purpose he invited Dutch peasants to extinction of the oyster. settle on the nearby island of Amager. Pota- toes took a long time coming to Denmark, These ideas have largely been abandoned. and the people who finally began to grow Others (e.g. Jennbert 1984; Fischer 2002:343– potatoes and teach this art to the Danes were 385) have proposed social mechanisms be- Germans who had been invited in the late hind the economic shift. Imported grain and 18th century to cultivate the Jutish moors. livestock may – like the shoe- last axes – have Returning to prehistory, it has been men- conferred status. If the leaders of farming tioned above that elements of LBK’s agricul- communities could host feasts with pork and tural technology spread very slowly north- beer, then they had a competitive edge on ward or westward to Holstein during the the hunters and fishers. But such a scenario 5th millennium cal BC. When EBK groups in is hard to imagine where great distances northern Europe acquired pottery technology separated the different populations and through contacts with farmers, it took about precluded frequent contact. 700 years for it to spread from Holstein to modern Denmark (Klassen 2004:111). Only The question of motives for autonomous two vessel types were concerned: pointed- cultural change has not been answered by base pots and pottery lamps. Neolithisation, the opponents of the immigration hypoth- however, entailed the transferral of an entire esis. This hypothesis can, however, offer an technological package: 1) agriculture, 2) stock explanation. Immigrant Neolithic popula- breeding, 3) new pottery with thin-walled tions placed the Mesolithic locals under and more durable funnel beakers in many stress. Certain important coastal Ertebølle specialised shapes, and 4) flintaxe grinding. settlements and the Jutish inland settlement There was also a cultural shift regarding 5) of Ringkloster were abandoned about the burial customs, and 6) widespread wetland time of neolithisation. This did most likely sacrifice of axes and pots containing food not occur by the inhabitants’ free will. Late and/or drink. Almost all of these elements Mesolithic culture could not survive the close were present where TRB replaced EBK, al- competitive encounter with Neolithic groups. though there are differences in how fast they This does not preclude the possibility of a appeared or have been documented. This voluntary shift in some localities. There are new knowledge or attitude could not have indications that traditional food sources, e.g. been acquired no matter how long there had the fishing that had supplied the staple food been trade and cultural contact. Whatever of EBK, played a continued important role may have been learned during trips to the into the Neolithic. Continent about 4100 cal BC would have been forgotten. The appearance of various

94 Figure 1. Left, two common axe types from the late Ertebølle Culture (EBK): a core axe and a ﬂake axe (c. 15 and 10 cm long, respectively). Top centre, a shoe-last axe, c. 15 cm. Below and to the right, two axe types characteristic of the early Funnel Beaker Culture (TRB): a pointed-butted and a thin-butted axe from the Oxie and Volling sub-cultures respectively. Their length vary from c. 15 to 40 cm. Reproduced after Jeg ser på oldsager, Politikens Forlag 1979.

Figur 1. Til venstre ses to hyppigt forekommende økseformer under sen ertebøllekultur: en kerneøkse og en skiveøkse (hhv. ca. 15 og 10 cm høje). I midten øverst den i teksten omtalte skolæstøkse, ca. 15 cm. Derunder samt til højre to typiske økseformer fra tidlig TRB: en spidsnakket og en tyndnakket økse fra hhv. Oxie- og Vollingkulturen. Deres længde varierer mellem c. 15 og 40 cm. (Gengivet efter Jeg ser på oldsager, Politikens Forlag 1979).

95 new technologies together with new cultural Klassen’s view of the two groups’ areas of elements in southern Scandinavia can only origin fit well with this conclusion. He be- be explained through the immigration of lieves that limited immigration took place farmers. when the Oxie group spread to southern Scandinavia, but does not say anything about Two subcultures in the Early Neolithic how the Volling group appeared. Consider- As mentioned above, there were two Early ing Volling’s likely area of origin toward the Neolithic culture groups in Denmark and English Channel, no pre- or Early Neolithic Scania, Oxie and Volling, that display differ- contacts with EBK would have been possible. ences both regarding material and ideational The only possible explanation for Volling’s culture. Oxie’s flint assemblage and pottery appearance is that the culture’s bearers im- design is closest to those of EBK, and its ori- migrated. They must have passed parts of the gins must according to Klassen (2004:323) Continent that had already been neolithised, be sought in Central Europe, more precisely apparently without leaving archaeological southern Lower Saxony or the Mittelelbe/ traces. The Neolithic groups that migrated Saale area. Volling, on the other hand, is into Scandinavia would thus have come according to Klassen similar to cultures in from both branches of the original wave the westernmost parts of Europe which had that brought the Neolithic to Europe 9000 not been influenced by the LBK. This goes years ago. for the round-based and highly decorated The Oxie group probably brought shoe-last pottery as well as the non-megalithic long- axes, many of which ended up as sacrifices barrows that are both found in the Paris in the bogs of the Danish Isles and Scania. Basin and England. The Volling group also 33 shoe-last axes of early types have been introduced flint mining. These two contem- found in the EBK area, only 4 of them (12%) poraneous groups with different culture in modern Denmark and Scania. 59 late shoe- co-existed on Zealand among other places last axes, produced up until c. 4000 cal BC, are (Koch 1998:181–185). known from the EBK area. 26 of them (44%) have been found in modern Denmark and Regardless of whether Klassen’s views of Scania. This marked change in the type’s dis- these two groups’ areas of origin are cor- tribution could be due to most of the 26 axes rect, the existence of two Early Neolithic having been brought by migrating farmers. subcultures in Southern Scandinavia cannot The axes often show signs of wear and repair. be explained without reference to immigra- They were thus probably not, as assumed by tion. Klassen among others, imported before 4000 Otherwise one part of late Mesolithic popu- cal BC as prestige objects into EBK. They were lations of Jutland, Zealand and Scania must most likely brought by migrating farmers have had long-standing contacts with one as agricultural implements, regardless of Continental Neolithic group, while another the fact that they were neither necessary part of the population in the three areas nor practical for this purpose on the stony had contacts with another Neolithic group. light moraine soils of southern Scandinavia. This is inconceivable. Immigration, however, The shoe-last axes are thus not evidence of offers a simple explanation: two farming intensive contacts with Continental farmers groups have immigrated into the Jutish pe- during the Late Mesolithic. ninsula: first Oxie, then a century or two later, Volling, which was probably the larger Bornholm and Gotland group. Volling expanded across Jutland and The island of Bornholm, located in the Baltic onward to the isles and Scania. It put pressure Sea c. 40 km south-east of Scania, was land- on Oxie, a main part of which had moved via locked southward in the Early Maglemosian Fehmarn to Lolland and onward to Zealand and became populated from the Continent. and Scania. Apart from the areas mentioned, Bornholm has many settlement sites from both groups are also present on Bornholm the interval 8300–6800 cal BC, but the finds and around Lake Mälaren. dwindle swiftly over time, and Bornholm

96 Figure 2. Two pottery types of the EBK. Left, a pointed- base pot, c. 30 cm high, above, a lamp c. 25 cm long.

Figur 2. To fra ertebøllekulturen kendte keramikformer, til venstre en spidsbundet lerkrukke, ca. 25 cm høj, øverst en lampe ca. 25 cm lang.

Two Early Neolithic TRB pots. Left, an almost undeco- To lerkrukker fra tidlig TRB, til venstre et næsten ude- rated flat-based funnel beaker, characteristic of the Oxie koreret tragtbæger med flad bund, karakteristisk for sub-culture. Right, a small decorated round-based funnel Oxiekulturen, til højre et lille rundbundet, dekoreret beaker, from the Volling sub-culture. Volling has many tragtbæger fra Vollingkulturen. Fra denne eksisterer der pottery types for different purposes. Reproduced after mange typer af krukker, tilpasset forskelligt brug. (Gen- Jeg ser på oldsager, Politikens Forlag 1979. givet efter Jeg ser på oldsager, Politikens Forlag 1979). became isolated by the rising sea level during Animal species often become extinct when this period (F.O. Nielsen 1996). Then follow isolated in small areas, and the same may 2500 years without any finds. Big game popu- happen to people. Game was dwindling and lations declined during this period (Aaris-Sø- the population may have been too small to rensen 1998:127–128; Vang Petersen 2001). change its subsistence strategy towards fish- Apparently the island became depopulated. ing and seal and porpoise hunting. They were

97 Three maps of settlement sites on Bornholm. There are no known settlement sites on Bornholm from 6800–4300 cal BC. Then the Ertebølle Culture reached the island. Reproduced after Forhistoriske interesser, Bornholms Amt 1996.

Tre kort over bopladser på Bornholm. Der er i de 2500 år fra 6800 f. Kr. indtil ertebøllekulturen nåede øen ca. 4300 f. Kr. ikke fundet bopladser på Bornholm. (Gengivet efter Forhistoriske interesser, Bornholms Amt 1996)

Southern Scandinavia

98 too few to develop their Maglemosian cul- den, apparently at about the same time as ture into something similar to the Kongemo- Bornholm was re-populated. sian or EBK. EBK appears on Bornholm about Bornholm and Gotland were neolithised by 4300 cal BC. A number of coastal sites are migration from mainland Sweden. It may be known, and they show no sign of earlier oc- discussed if this tells us with certainty that cupation. This re-population seems to have migration played the same role in Denmark, taken place through migration from Scania. Scania, the Lake Mälaren area and the Oslo TRB reached the island c. 500 years later, Fiord. A model where Neolithic cultures without doubt from Scania as well, as shown spread by different mechanisms to differ- by similar finds in the two areas. ent parts of the area should however in my opinion require support in the archaeological If TRB developed locally on Bornholm record, which is lacking. through the impact of external influences, these contacts must have touched eastern Scania and not the southern shore of the How did Neolithic culture spread through Baltic. However, archaeology shows that Scandinavia? Bornholm did not have close contacts with As discussed above, datings from the time of Scania during the Late Mesolithic. Bornholm’s the economic shift c. 4000 cal BC in southern geology only offered small round flint nod- Scandinavia suffer from various uncertainties. ules that only allowed the production of It appears fairly certain, however, that Neo- small objects such as arrowheads. Larger lithic culture took only c. 200 years to spread flint objects are extremely rare in Bornholm’s from Holstein through Denmark to Scania EBK. If contact had been intensive, flint for and Bornholm, and on to Lake Mälaren, e.g. axes would have been imported from Gotland and the Oslo Fiord. Large numbers Scania. Such an importation was not estab- of people passed straits and open sea with lished until the Neolithic, when it delivered their livestock, seed corn and implements, a considerable volume of flint, as seen by and it appears unlikely that this could have numerous finds of larger implements from been done with log dugouts, the only boats this period. available at the time. During EBK’s final phase on Bornholm, The rapid spread was apparently made Neolithic cultures were newly established possible by climate change that has been in Scania. Considering the clearly limited demonstrated for the period in question. contacts, it appears that it would have been Post-glacial warming since c. 13000 cal BC impossible for the fishers of Bornholm to had not proceeded linearly. Temperature acquire the knowledge necessary to switch fluctuated until c. 9000 cal BC. Then mean economies on their own. Diffusion without July temperatures rose from c. 14°C to c. 19°C migration is inconceivable, so Bornholm must in the late 5th millennium cal BC (the final have become neolithised through migration centuries of the EBK). This was 2–3° warmer from Scania. than the mean temperature in Denmark A corresponding argument can be made for during the past 40 years. Then, however, Gotland, an island located c. 80 km from climate changed again. July temperatures the Swedish mainland. It became populated fell slightly, and at the same time climate c. 7500 cal BC, probably by people passing became more continental with colder winters across the ice in wintertime. Northern Scan- (cf. Berglund 1991:69). dinavia was still partly covered by the inland The melting of the inland ice and the sub- ice at this time. People subsisted on fishing sequent rise of the land led to dramatic and seal and porpoise hunting, as the only changes in the contact of the Baltic Sea with larger land mammals on Gotland at the time the Atlantic. This affected salinity. The Baltic were foxes and hares. A period of about was a fresh water lake (the Ancylus Lake) 1000 years until c. 4300 cal BC shows no sign from c. 8400–6500 cal BC. Its drainage point of habitation. Then fishers and gatherers into Kattegat varied. Then the modern straits reappeared, probably from mainland Swe- of the Belts and Øresund formed, permit-

99 ting the entry of salt water into the Baltic scholars have thus received the erroneous im- basin. About 4000 cal BC, salinity was once pression that neolithisation hit the southern again reduced in Kattegat and the Baltic part of Scandinavia explosively (cf. Hallgren due to rising land and lessening tidal effects. 1996; Pettersson 1999; Price 2000:293). On Together with the colder winters, this must the basis of calibrated radiocarbon dates, have meant that straits and sea often froze Fischer (2002:355–356) concluded that the over. Such was the case, for instance, in AD year 3950 cal BC marks both the abrupt end 1657–1658, which permitted King Carolus X of the EBK and the instant onslaught of the of Sweden to invade Denmark on horseback. TRB. This view appears weakly founded. This happened during the ”little ice age” from c. AD 1550–1700, a period with a mean The idea that the Neolithic appeared all over temperature only 0.5° lower than current southern Scandinavia at the same time has temperatures (Aaris-Sørensen 1998:202). led many scholars astray. Hallgren (1996:18) In this way, the migrating Neolithic farmers inferred that there could not have been any would have been able to pass the Belts and immigration. Immigration would according Øresund in the wintertime to reach the Isles to his view have taken longer as the area and Scania, and then continue to Bornholm involved is so large. Other Swedish archaeolo- and along the icebound eastern coast of gists have also apparently been inﬂuenced by Sweden north to Lake Mälaren and Gotland, the seemingly rapid spread of neolithisation as well as along the western coast to the across Scandinavia. Oslo Fiord. One example is Kihlstedt (1997:122) who did This also explains the varying communica- not however entirely rule out migration as an tions between mainland Sweden and Born- explanation of the spread from Scania to lake holm and Gotland respectively. Only after Mälaren. It is not apparent, however, that the climate change did the islands have other suggested explanations for a spread stable connections with the mainland. The that is erroneously seen as instant are any re-population of the two islands c. 4300 cal better: e.g. frequent trade contacts in the BC may be connected with a short phase of 5th millennium (the availability model) or lowered temperatures that allowed travel cultural community and exchange of spouses. across the ice. Such a short cold spell occurred These mechanisms would also have required for instance during World War II. time.

Malmer (2002:176) has pointed out that, if How long did the spread of Neolithic cul- there were a consensus that the spread of a ture take from Holstein to Lake Mälaren? certain culture or economy took place from Many radiocarbon dates for Early Neolithic a certain direction – in this case, from the ﬁnds and contexts calibrate to 3950 or 3960 south – then we might choose among the cal BC. This applies to Jutland, Zealand, Scania, probability maxima along a wiggle in an the Lake Mälaren area and Norway (cf. e.g. OxCal graph, taking the geographic location Price 2000:271, 284, 287; Fischer 2002:346, of the site under study into consideration. If 358). The explanation should probably be Malmer’s idea is accepted, one would choose sought in the widespread use of the CALIB the earliest intersection with the calibration computer program from Seattle. CALIB often curve of an uncalibrated date for the neoli- points out a single year as the calibrated thisation of Denmark (and possibly Scania). date even when the probability distribu- Later intersection points would be more tion behind it is skewed. This also happens likely for datings of TRB’s arrival around Lake in cases where wiggles in the calibration Mälaren or the Oslo Fiord. This would allow curve produce (typically) three equally valid TRB c. 200 years to travel from Holstein to calibrated dates for one date bp. This is the Lake Mälaren and the Oslo Fiord. case for datings from c. 3120–3020 bc where the range of possible calibrated dates for one uncalibrated date is typically 150 years. Many

100 What was the scale of the immigration? – 1991: Norsminde. A Køkkenmødding with In recent years a number of archaeologists Late Mesolithic and Early Neolithic Occu- have conceded that a certain amount of im- pation. Journal of Danish Archaeology 8, migration into Scandinavia may have taken 1989. place in the Early Neolithic. They have, how- – 1993a: Bjørnsholm. A Stratified Køkken- ever, insisted that this happened only on a mødding on the Central Limfjord, North small scale (Price 2000:293; Fischer 2002:381; Jutland. Journal of Danish Archaeology 10, Malmer 2002:178; Klassen 2004:338), al- 1991. pp. 59-96. though no arguments for the small scale of – 1993b: Kystens Bopladser. I: S. Hvass og B. immigration have been put forward. Storgaard (red.): Da klinger i muld… 25 års This discussion is difficult without a shared arkæologi i Danmark, 1993. pp. 65-68. definition of how large a small scale is. The – 1998: Ringkloster. Ertebølle trappers and fact that TRB spread across a very large area wild boar hunters in early eastern Jutland. in a rather short time indicates that the im- Journal of Danish Archaeology 12, 1994–95. migrants were fairly numerous. It should also pp. 13-59. be taken into account that there seems to – 2000: Fisker og bonde ved Visborg. I S. Hvass have been an abrupt shift from EBK to TRB og det arkæologiske Nævn (red.) Vor skjulte at most coastal Danish sites. Occupation of kulturarv. pp. 42-43. some Danish EBK sites even ends entirely Andersen S.H. & Storgård, B. (red.) 1987: around the time of the appearance of TRB’s Ertebølle revisited. Journal of Danish Archa- Oxie sub-culture. Immigration thus seems to eology 5, 1986. pp. 31-61. have been so considerable already at this Bellwood, P. S. 2004: First Farmers. The origin early date that it did not confine itself to the of agricultural societies. Blackwell publish- inland, where farmers could have established ing. themselves without any great friction with Berglund, B.E. (ed.). 1991. The cultural land- the original Mesolithic population. It appar- scape during 6000 years in Southern Sweden. ently put the coast dwellers under consider- The Ystad project. Ecological Bulletin 41. able pressure. Copenhagen. Fischer, A. 1982: Trade in Danubian Shaft- Therefore it seems likely that immigrants Hole Axes and the Introduction of Neolithic to Denmark – even disregarding the later Economy in Denmark. Journal of Danish Ar- immigration connected to the spread of the chaeology 1, 1982, pp. 7-12. Volling sub-culture – outnumbered the lo- – 2002 “Food for feasting” in Fischer and cal Mesolithic population. Immigration to Kristiansen, (ed.) 2002. The neolithisation of Sweden, where neolithisation appears to Denmark, 150 years of debate. Sheffield. have been more gradual, may have been less Fischer, A. & Kristiansen, K. (ed.) 2002. The massive than in Denmark, seen in relation to neolithisation of Denmark, 150 years of de- the local Mesolithic population numbers. bate. Sheffield. Hallgren, F. 1996, Sociale territorier och ex- Translated from Danish by ogamirelationer i senmesolitisk tid, etc. Tor Martin Rundkvist, PhD. 28: 3-27 Hartz, S. et al. 2002. Coastal Farmers - the Niels V. Skak-Nielsen, neolithisation of northernmost Germany. In Gentoftegade 2.tv., Fischer and Kristiansen (ed) 2002. The neo- DK-2820 Gentofte lithisation of Denmark, 150 years of debate. [email protected] Sheffield. Hesjedal, A. et al. 1996. Arkæologi på Slet- tnes. Dokumentasjon av 11.000 års bosetning. Bibliography Tromsø museums skrifter 26. Tromsø. Andersen, S. H. 1988: Mesolithic dug-outs Hvass, S. & Storgård, B (red.) 1993. Da klinger and paddles from Tybrind Vig, Denmark. Acta i muld ... 25 års arkæologi i Danmark. Det kgl. Archaeologica 57, 1986, København.

101 nordiske oldskriftsselskab & Jysk arkæologisk – 1993: Høje med træbyggede grave. I: S. selskab. København. Hvass & B. Storgaard (red.): Da klinger i Hvass, S. & det arkæologiske Nævn (red.) muld… 25 års arkæologi i Danmark pp. 96- 2000. Vor skjulte kulturarv. København. 99. Jennbert, K. 1984: Den produktiva gåvan. Malmer, M. P. 2002. The Neolithic of South Tradition och innovation i Sydskandinavien Sweden. KVHAA. Stockholm. för omkring 5 300 år sedan. Acta Archaeo- Midgley, M.S. 1992: TRB Culture. The First logica. Lundensia, no. 16, Lund. Farmers of the North European Plain. Edin- Jensen, J. 2001: Danmarks stenalder, Gylden- burgh. dal. Copenhagen. Nielsen, F. O. 1988: Bornholms bebyggelse Karsten, P. 1994: Att kasta yxan i sjön. Acta i yngre stenalder. Fra Bornholms Museum Archaeologica. Lundensia, no. 23, Lund. 1987-88. Rønne. Kihlstedt, B. et al. 1997: Östra Mellansverige – 1996: Forhistoriske interesser. Bornholms + Neolitiseringen som et socialt fenomen. I Amt. Rønne. Larsson, M. & Olsson, E. (ed.) Regionalt och Nielsen, P. O. 1987: The Beginning of the Interregionalt. Stockholm. Neolithic - Assimilation or Complex Change? Klassen, L. 1999: The debate on the Meso- Journal of Danish Archaeology 5, 1986. pp. lithic-Neolithic transition in the western Bal- 240-243. tic: a central European perspective. Journal Persson, P. 1999: Neolitikums början. Göte- of Danish Archaeology 13, 1996-97. borg. – 2002. The Ertebølle Culture and Neolithic Petersson, H. 1999: Where did all the farmers continental Europe. In: Fischer & Kristiansen come from? Journal of Danish Archaeology (eds). The neolithisation of Denmark, 150 13. years of debate. Shefﬁeld. Price, T. D. (ed.). 2000. Europe’s First Farmers. – 2004. Jade und Kupfer. Jutland Archaeo- Cambridge. logical Society, vol 47, Århus. Price, T. D. et al. 2001. Prehistoric human mi- Kristiansen, K. 1988: Det danske landbrugs gration in the Linearbandkeramik of Central historie. Oldtid til 500 f.Kr. I C. Bjørn (red.). Europe. Antiquity vol. 75, nr. 289. Det danske landbrugs historie. Bind 1, Renfrew, Colin 1987: Archaeology and Langu- Odense. age. The problem of Indo-European Origins. Koch, E. 1998: Neolithic Bog Pots from Zea- London. land, Møn, Lolland-Falster. København. Rowley Conwy, P. & Zvelebil, M. 1986. For- Larsson, M. 1987: Neolithization in Scania – A agers and farmers in Atlantic Europe. In: Funnel Beaker Perspective. Journal of Danish M. Zvelebil (ed.), Hunters in transition, pp. Archaeology 5, 1986. pp. 244-247 67-93. Cambridge University Press. Larsson, M. & Olsson, E. (red.) 1997 Regionalt Skak-Nielsen, N.V. 2003. Hvordan kom bonde- och Interregionalt. Stenaldersundersökningar bruget til Sydskandinavien? + Addendum i Syd- och Mellansverige. Riksantikvarieäm- hertil. Fornvännen 98 (2003:1 & 2 ). betet: Skrifter nr. 23. Stockholm – 2004. Hvordan kom bondestenalderen til Dan- Lindqvist, Ch. 1994. Fångstfolkets bilder. mark? Fortid og Nutid 2004:1. København. En studie av de nordfennoskandiska kustan- Stafford, M. 1999. From forager to Farmers knutna jägarhällristningarna. Theses and in Flint. Aarhus. papers in archaeology. New series A5. Uni- Tilley, C. 1996 An ethnography of the Neo- versity of Stockholm. lithic. Cambridge. Lindqvist, Ch. & Possnert, G. 1999. The First Vang Petersen, P. 2001. Grisby – en fang- Seal Hunter Families on Gotland. Current stboplads fra ertebølletid på Bornholm. I Swedish Archaeology 7. Stockholm. Lass Jensen, O. (ed.). Danmarks Jægerste- Madsen, T. 1987: Where did all the hunters nalder. Status og Perspektiver. Hørsholm go? - an Assessment of an Epoch-Making Epi- Egns Museum. sode in Danish Prehistory. Journal of Danish Aaris-Sørensen, K. 1998. Danmarks forhisto- Archaeology 5, 1986. pp. 229-239. riske dyreverden. Gyldendal. Copenhagen.

102