Debate Prehistoric Migrations - the other cases (Hachmann 1970, Clarke 1968, ch. 9), just as the ethnographic record showed no clear pattern (Hodder 1978b). Case of the Single Grave and It seemed increasingly difficult to establish reliable criteria that could be used more generally (e.g. Crossland & Birchall 1974, Corded Ware Cultures Thompson 1957, Arutjunov and Chazanov 1981, Rouse 1986). On the other hand it could be shown that an internal frame­ By KRISTIAN KRISTIANSEN work of social and economic change often accounted more convincingly for the evidence as part of an autonomous deve­ lopment (summarized in Renfrew 1973 and 1979). Functional adjustments to various forms of social and ecological stress in MIGRATION IN RETROSPECT combination with international information exchange adapted to local needs were seen as regulating factors (e.g. Renfrew and Diffusion and migrations are phenomena on a continuous Shennan 1982, Bintliff 1985). Soon this relegated the concepts scale of cultural and social interaction and change. Through­ of migration and diffusion from the realm of serious archaeo­ out the 1960's and 70's such studies concentrated upon devel­ logical discussion within the new archaeology of the 1960's and oping our knowledge of the basic forms of such interaction - 1970's. Today it is implicitly accepted that migrations played no from reciprocal exchange over elite exchange/prestige goods significant role in the course of European prehistory (e.g. exchange to trade (Earle & Ericsson 1977, Renfrew 1975). It Champion et al. 1984), also demonstrated in the latest work of has, however, become increasingly clear that a prerequisite for Colin Renfrew (1987). Some point of critique should be raised such studies is a better understanding of the social formation against this approach: and constitution of culture as a spatial phenomenon. From 1. Modern archaeology has convincingly demonstrated that acknowledging the complexity of the problem (Hodder 1978a) material culture is complex and rarely reducible to an overlap­ we have during the 1980's seen a increasing number of studies ping pattern of cultural traits (Clarke 1968 fig. 58), a basic no­ trying to delineate some of the mechanisms by which material tion behind the traditional concept of culture, language and culture is constituted and maintained as part of social and po­ ethnicity. However, it is still implicitly believed that a migration litical strategies (e.g. Hodder 1982a & b). presupposes an unchanged geographical movement of recur­ Since diffusion and migrations were among the most criti- rent cultural traits, otherwise it is refuted (Shennan 1978). zed explanatory concepts of so-called traditional archaeology Thus the new archaeology has, as a paradox, maintained the modern archaeology has not yet come to terms with them traditional notion of culture as a one dimensional phenome­ either in archaeological or in theoretical terms. This paper is a non in its critique of migrations. One reason for this is of course preliminary attempt to incorporate the geographical move­ the refusal to take population movements seriously, since they ment of social groups into the conceptual and explanatory were not considered relevant to explaining social change framework of archaeology.1 How do we delineate various types Therefore the concept was not dealt with in a systematic way. of migrations against such phenomena as elite exchange, trade But I believe there is more to it. and marriage alliances? And how do we account for such phe­ 2. Just as the old parallelism between cultural change and nomena in structural and evolutionary terms? Before answe­ migrations was rooted in a modern notion of national and po­ ring these questions it will be useful to discuss the background litical history, cultures and migrations replacing nations and to the present situation in more detail. battles, so it can be argued that the prevailing parallelism be­ Throughout the 1960's and 70's a number of studies demon­ tween social change and peaceful internal development is strated the archaeological inconsistencies and inadequate theo­ rooted in post-war decolonization and the development of mo­ retical status of prehistoric migrations and diffusion (Adams dern middle class welfare society, international information ex­ 1968, Binford 1968, Clark 1966, Myhre & Myhre 1972, Renfrew change and internal social change substituting for internatio­ 1973 & 1979). Although a number of studies that combined his­ nal cooperation (United Nations, EC etc.) and social reforms. torical and archaeological sources could demonstrate convin­ Culture, ethnicity and migrations were thus seen as linked to cing regularities between ethnic groups and material culture the political ideology that led to the disasters of two world wars. (Hachmann, Kosasack & Kuhn 1962), this proved impossible in (Jensen 1988, Klejn 1974). A new theoretical framework was 212 therefore needed that was in accordance with the political ideo­ on demonstrating internal change from the TRB to the SGC logy after World War Two. It became one of evolution, progress (e.g. Ebbesen 1980, Kristiansen 1982, Malmros 1979) in combi­ and peaceful internal development. I propose that these nation with international information exchange of new ritual changes in ideological climate of the present are important in and social value systems. Since such explanations had taken order to understand some of the reluctance of modern precedence in explaining the Corded Ware/Battle Axe Com­ archaeology' to deal with the traditional concepts of culture, plex (Maimer 1962, Haussler 1963 & 76, Neustupny 1969), it ethnicity and migrations. This, however, had some serious con- was difficult to maintain a different explanation for the Danish cequences. case. No systematic attempt was ever made, however, to refute 3. A theoretical and methodological framework without de­or confirm the migration hypothesis. With few exceptions (Da­ vices for identifying and interpreting the movement of social vidsen 1975 and 1978, Jorgensen 1977) there simply occurred and/or ethnic groups, normally labelled under the general a drift in approach that was never sustained by systematic term migration, is unlikely to make convincing progress in research (Becker 1981 for a summary). Let us therefore in other Fields of social and cultural interaction. To exclude one some detail consider the criteria employed to support the two phenomenon of social and cultural change in favour of others hypotheses. distorts our general ability to identify and explain such change. A framework of social change should thus include both conflict The migration hypothesis: and harmony, migrations and information exchange. Migrati­ ons may both be a result of and result in social and economic 1. The SGC appears at once and fully developed. There are no disruption, including geographical displacement and warfare. links to the existing TRB culture. SGC differs from TRB in The inclusion of the study of migrations into modern archaeo­ terms of material culture, technology (pottery, flint), religion, logy, however, makes it necessary to make certain theoretical and, as we now know, social organisation and subsistence (see claims. First it should be made clear that any such study should discussion below for references). be contextualized, culturally and structurally. There exist no 2. Its primary area of settlement, the more marginal soils in universal categories that allow the identification and explanati­ central, western and southern Jutland, mostly lies outside the on of such phenomena. Second: any such study must be eval­ settlement areas of the TRB culture. The two cultures are thus uated against the historical background preceeding it. Only in in the earliest phase mutually exclusive (Davidsen 1975, fig. 7), this way can changes be identified and explained. Third: a mi­ with a brief period of chronological overlap at the peripheries gration, of whatever kind, is always a sympton, not a primary of expansion (Damm 1989). cause, and so it has to be explained within a broader framework 3. Where geographical overlap with the TRB culture occurs of social organisation, contradiction and change in the initial phase it represents a break of cultural continuity, In a recent paper I have tried to take this into account on a the TRB comes to a complete stop, and is replaced by the SGC larger scale of social transformation in temperate Eurasia (Kri­ (Rostholm 1982, j0rgensen 1977 & 1985). stiansen 1989). In the following I shall concentrate on a case 4. There is virtually no evidence of contact between the SGC study, that of the Single Grave Culture, or Battle Axe Culture, and still existing TRB culture groups in eastern Denmark. Am­ in Jutland, which is part of the larger complex of Corded Ware ber, controlled by the SGC, thus disappears from the TRB cul­ cultures that spread throughout Europe during the early third ture, just as good flint and its technology, controlled by the millenium. The objective is to create a more systematic ar­ TRB, is not available to the SGC (Ebbesen 1986). Ebbesen con­ chaeological basis for analyzing and evaluating the question of cludes his analysis in the following statement: ‘Thus during MN migration and social change. В (the time of the SGC) there existed a distributional, and pro­ bably also a communicational barrier between the classic SGC regions in middle and central Jutland and the rest of the THE SINGLE GRAVE CULTURE - A CASE STUDY country” (Ebbesen 1986, 37 f.). 5. The subsequent stages indicate a continous expansion of In recent years the conception of the origin of the Single Grave settlement; that is, a slowed down continuation of migration in­ Culture has changed according to the general shift in explana­ to previously settled regions of the TRB culture, creating a tory framework within archaeology. Earlier scholars such as mixed culture (Skaarup 1986, Andersen 1986). Sophus Muller (1898) and P. V. Glob (1944) saw this culture as It should be noted that these findings are based upon one of representing a migration into Denmark of Indo-European the most complete and representative archaeological materials speaking peoples, bringing with them a new, dominant culture in archaeology, since the SGC was systematically excavated in a that gradually took over and subordinated the peaceful mega- large campaign in the late 19th century, later followed by nu­ lithic people. Out of this evolved the ranked Bronze Age soci­ merous excavations. The material has been systematically ana­ ety. With the advent of Cl 4 it became clear that the Single Gra­ lysed by Glob (1944), Struve (1955) and the preceding period ve Culture succeeded the TRB or Megalithic Culture in Jutland of the TRB culture in Jutland by Davidsen (1978). Local in (Malmros and Tauber 1975), whereas the latter still lived on in depth studies confirms this picture, both regional settlement the Danish islands in modified form (Davidsen 1980). In com­ surveys (Mathiassen 1948. Skamby 1984) and local excavation bination with the new trends in archaeology research focused programs, although we lack a modern treatment of the numer­ 213 ous finds since Globs work. Although archaeological formation very likely that the TRB culture could supply the numbers of processes could be responsible for some of the variation be­ people indicated by the archaeological and palaeobotanical re­ tween Jutland and Eastern Denmark (Maimer 1986), recent cord, showing that most of central and westernjutland was sett­ research has confirmed that regional and chronological diffe­ led by the SGC, except if there still existed a local population rences between the SGC and the TRB are to be considered real of hunters and fishers. The evidence does not point towards and representative for eastern Denmark (Ebbesen 1986, An­ surviving hunter/fishers in Denmark (Andersen and Sterum dersen 1986, Skaarup 1985). 1970-71). On the other hand it is known that demographic growth can be rapid, and we know too little about TRB settle­ ment continuity/discontinuity in Eastern Jutland/Denmark. The autonomous hypothesis: Mapping of single finds, however, suggest some setdement con­ It follows from the above observations (1-4) that it is impossible tinuity (Nielsen 1977: fig. 14, Ebbesen 1986), which does not to point out traits that indicate cultural continuity between the support local migration and cultural change on a larger scale TRB and the SGC. One can, however, point out a number of within Jutland. changes within the TRB culture that may account for the readi­ Although the TRB culture had already transformed the fo­ ness of the final TRB to adopt a new social and cultural organi­ rest into pastures in some regions in Judand (Odgaard 1985), sation in Judand. They are: especially in eastern Denmark (Andersen 1985, Andersen et al. 1. A gradual change in ecology and economy in some regi­ 1984), the SGC is characterized by a major clearance and burn­ ons towards open pastures and husbandry, the dominant sub­ ing horizon throghout Jutland (a real “landnam” much more sistence strategy of the SGC (Davidsen 1978: 140 ff., Madsen extensive than the earlier TRB “landnam”), whose main pur­ 1982, fig. 17). pose was the creation of heathland or pastures for large herds 2. A local change in burial customs in Jutland towards single (Odgaard 1985 & 1987, Andersen in press). No agricultural in­ burials in stone packed flat graves, although this was still related dicators occur at this stage, and no house structures have yet to the megalith and different from the subsequent burial been identified.2 In the later stages houses are small and pardy customs of the SGC in barrows (Jorgensen 1977). subterranean, occurring in small clusters of two or three (Hvass 3. Pressure from expanding coastal fishers and hunters, the 1986). Some agriculture was practiced, however, although Pitted Ware Culture (Becker 1980), from Sweden and the grain impressions on pottery are much less frequent than in the Baltic, leading to some changes in material culture (e.g. potte­ TRB (Rostholm 1986). Recent evidence also suggest that the ry), and economy (e.g. hunting). This might be seen as re­ SGC cultivated barley only, in opposition to the TRB which pre­ flecting a crisis of the traditional farming communities, in com­ ferred wheat (Robinson & Kempfner 1987, Hedeager & Kristi­ bination with climatic change (Kristiansen 1982: 260, Hede- ansen 1987, 76 ff. for a recent summary). ager& Kristiansen 1988: 71 ff.). If one accepts the autonomous hypothesis it also has to be According to this scenario, expansion of settlement onto the explained why there was no contact between the mother group marginal lands in Jutland, as a result of internal crisis of the (the TRB culture) and its offspring (the SGC), except if one en­ TRB culture, led to radical social and cultural changes. Thus visage a revolution, stimulated by the new expanding SGC ideo­ the SGC is regarded as a social and ecological adaptation to logy, followed by warfare and local migrations. This presuppo­ marginal environments, just as the Pitted Ware Culture is regar­ ses a build up of local contradictions and political organization ded as an adaptation to the Swedish and Danish coasts. on a scale I consider unlikely, although contradictions within This internal framework, however, fails to account for a the TRB groups were presumably part of the process of social number of features that remains unexplained or only partly ex­ transformation in several regions in Europe, where the two cul­ plained. Anthropology teaches us that significant cultural chan­ tures merged. In Judand, however, it is obviously more likely ges may occur as a response to external and internal crisis, e.g. that such a situation was the result of an immigrating people religious movements, although most cases relate to the effects which the TRB culture resisted by all means. Such a cultural of western imperialism (e.g. Wallace 1970). In the case of the barrier corresponds well to a situation on internal stress and re­ Single Grave Culture the change was complete - within a gene- sistance between two ethnic groups, as proposed by Hodder radon or two a new and mature cultural, religious and social (1979, plus examples in Hodder 1982). During the process of framework was in place.To imagine that this should have hap­ expansion, however, it seems likely that many TRB people in pened as a internal transformation, from a culture that in most Judand “convened" to the SGC ethnicity, since ethnicity is a cul­ respects was quite different, leaving no traces of the former cul­ tural code that can be adopted through socialization, e.g. mar­ ture (it should be stressed, once more, that the change is not riage alliances - or force (Damm unpublished). The social or­ only religious, but includes all major aspects of social and eco­ ganisation of the SGC was geared to expansion (Sherratt 1981), nomic life), demands support from the material evidence of by establishing new settlements through alliances and cultural the late TRB Culture in terms of demography (population pres­ inclusion, supplemented bv warfare, to secure domination, sure), abandonment of TRB setdements in Eastern Denmark, much in the same manner as described for tribal pastoralists in and some technological and cultural continuity, at least at the .Africa (Sahlins 1961. Bonte 1977). Milisaurus and Kruk have re­ level of cultural relicts. This support is difficult to mobilize, cently through detailed research of a micro-region in Poland looking at the published evidence. First of all it does not seem come to conclusions that would seem to support such a scena- 214 no (Milisaurus and Kruk 1989). In fact even researchers who Corded Ware complex. This makes comparisons difficult. Final­ reject the migration hyphothesis, agree that the SGC and Batde ly, no understanding of the origin of local groups is likely to ma­ Axe cultures were based upon an ideology of hierarchy, warfare terialize before the problems of the genesis of the whole Cor­ and domination (Maimer 1989). How such an ideology should ded Ware complex reaches a more mature stage in terms of be accepted peacefully by an alien culture, with whom there theoretical framework and archaeological analysis. Little has had been no previous contacts remains the paradox. happened in these respects since Mats Maimer's analytical de­ Also the continous eastward expansion over time of the SGC velopments (1962) and David Clarke's methodological propo­ into areas of TRB culture, leading to a number of cultural chan­ sals (Clarke 1968: Table II). At present two conflicting “models” ges (Skaarup 1986, Andersen 1986), apparently through both are at hand, basing themselves upon very different perspectives warfare and alliances, would seem to be a more likely scenario of cultural change, one giving priority to internal forces,- for a still expanding foreign people. another to external forces. They have been summarized most To this can be added that the cultural resistance or oppositi­ coherently in recent works by respectively Steve Shennan (1986 on between the original core areas of the SGC in Jutland, and a & b, critique Maimer 1989: 8) and Maria Gimbutas (1979, the rest of Denmark, continued to manifest itself very clearly in 1980 & 1986, critique Hausler 1985).4 the archaeological record until 1500 B.C., that is through more Shennan sees the changes as an interaction between chan­ than 1000 years, when the mature Nordic Bronze Age Culture ging ecological conditions of production and interregional ex­ finally integrated it within its framework (Kristiansen 1987). change of corresponding new social and religious value sy­ This adds a significant historical dimension to the migration hy­ stems, whereas Gimbutas rather sees changes as caused prima­ pothesis, since such resistant social and cultural traditions are rily by so-called Kurgan steppe pastoralists intruding into most probably to be seen as a result of ethnicity, and perhaps Eastern and Central Europe in a number of waves leading to so­ also a different language at first. cial and economic transformations. Shennan argues that chan­ Having discussed the two alternative hypotheses there is litt­ ges were peaceful, since dispersed settlements replaced more le left in support of a “pure” autonomous hypothesis. The evi­ centralized and fortified settlements (also Starling 1983), dence is, as it stands today, rather conclusive: the case of the whereas Gimbutas argues, on the basis of axes, new bow and ar­ SGC in Jutland must be considered to represent a classic ex­ row techniques and horse riding, that it was one of conflict. ample of a migrating, tribal people, settling within a very short Both agree that warfare must have taken on a new character period of time in a new, sparsely populated environment, large­ and that changes in social organisation were decisive. ly defined by resistance from existing TRB settlements. They Model 1 does not account for various types of population belonged to the Corded Ware/Battle Axe cultural complex, movements since they are a priori denied, whereas model 2 uses and show greatest familarity with similar groups stretching migrations as an explanation, instead of trying to explain why through Northern Germany (Schl.-Holstein) to the Nether­ they should occur. It is apparently taken for granted that Kur­ lands (van der Waals 1965), local groups in Switzerland gan people were expansive, but this needs qualification.5 In (Strahm 1971) and the Baltic/Poland (Kilian 1955, Machnik much the same way Shennan takes for granted that social inter­ 1981b, Wyszomirska 1989), with more remote links to the low­ action was the prime integrating mechanism. That also needs lands north of the Carpathians and the forest/steppe zone of qualification. None of these frameworks are thus fully satisfac­ the Pontic region (Machnik 1981a: Taf. 1, p. 281, Rulf 1981). tory, although Shennan's model is the more acceptable, since This of course is not the final word about the Single Grave it takes into account and tries to explain the actual processes of Culture. However, the exercise of presenting and testing the change from a theoretical perspective. A major critique to be two traditional alternative hypotheses has hopefully served its raised against Gimbutas' approach is that it has not responded heuristic purpose - to establish a more well defined and well ar­ to the theoretical critique of its prewar ethnic/migratorv gued platform for future discussions. The burden of falsificati­ framework, nor defined the conditions to be met in order to on now lies on the shoulders of supporters of the autonomous identify various types of migration and acculturation. There­ hypothesis. Having left the middle ground between the two fore, although some of the general historical trends may be cor­ alternatives rather open, I expect this to be more fully explored rect, acceptable theoretical and methodological underpin­ in the future. nings are lacking. Both Shennan's and Gimbutas' interpretations are models, or explanatory frameworks, trying to account for the historical THE CORDED WARE CULTURE - A REVIEW and social processes at work in general terms. To proceed from that we need to develop and apply such models in the working How do these findings relate to the European Corded Ware out of more specific case studies that take into account the Culture of which they are part?3 It is difficult to point out any whole variety of evidence, not only burials types or pottery, such obvious local parent group to the SGR Culture. What makes the asGallay (1981), Strahm (1981) or Machnik (1981) (fig. 1). Du­ problem even more intriguing, and interesting, is the fact that ring the prevailing discussion of the origin of the Corded Ware the SGC in Jutland is among the few regions where virgin sett­ Culture opponents and proponents of the migration hypthesis lement took place. Thus the original migrating cultural com­ have relied upon empirical studies of material culture without plex is intact, in opposition to most other local groups of the paying due consideration to the cultural and structural frame- 215 1@¥8І 1 lews! 2 level 3 level 4 hypothesis A hypothesis В hypothesis C Fig. 1. Explanatory models for the evolution of Bronze Age society in the Alpine region. The components of the models may interact simultaneously in various combinations through time and space (after Gallay 1981). work within which it operated. Proponents have often relied on 1980). But it is generally agreed that the CWC spread in a select list of traits, focusing only on points of similarity, but ig­ obedience to local conditions and exhibits a large variety both noring the local cultural context. Opponents, on the other in terms of the actual processes of expansion and in terms of hand, have focused on points of variation and have therefore cultural mix. This leads on to a consideration of the impact of been so eager to take all aspects of material culture into ac­ Kurgan traditions of the steppe and forest-steppe regions in the count that they have dissolved it into too many components, re­ actual formation of the Corded Ware and Batde Axe cultures. gardless of their internal social and cultural meaning (discussi­ Kurgan influences: Although both pottery and batde axes of on in Klejn 1969, Clarke 1968, 287 ff.). the CWC may be given a Central European origin, it also seems In order to throw some more light on the nature of change, rather obvious that this does not account for a number of and the present situation of research, I shall briefly discuss distinct features in burial ritual. Here a Kurgan origin is sdll some of the factors considered to be significant. most likely, although Hausler in his works maintain that the Genesis: Today most researchers agree that the genesis of the CWC of Cen tral Europe and the Ochergrabkultur, by some also Corded Ware complex is, for the major part, to be sought in called the Pit Grave,Jamna or Kurgan Culture, stretching from Central Europe, rooted in large scale economic and social the Volga to Hungary, represent two different cultural com­ changes, with Baden and Globular Amphorae Cultures playing plexes (Hausler 1963, 1967,1974 8c 1976). Such a distinction is amajorrole (Sochacki 1980, Rruk 1980, summarized in Sherrat obviously dependent upon definitions, but it seems to be gene­ 1981). These major economic changes and their continuity in rally agreed that the pastoral farmers of the Pit Grave or Ochre the Corded Ware Cultures were already summarized 20 years Grave Culture proper did not expand beyond the river Theiss ago by Neustupny (1968). They created a necessary economic in Hungary (Ecsedy 1979). It also seems clear, however, that it and social background, although they were also subject to ex­ is exacdy this mixture between Kurgan burial ritual and Corded ternal influences throughout their development, some of them Ware material culture that produces the classic package of the much discussed, e.g. Gimbutas' Kurgan wave 1 and 2. Thus the Battle Axe/Single Grave Culture in Northern Eurasia, or some Globular .Amphora Culture has an eastern branch, defined by of the classic early Corded Ware groups, as pointed out by many pottery, which cannot be derived from its western branch scholars (see especially Struwe 1955) .fi And it has not. in mv opi­ (Nortman 1985, in opposition to Sulimirski 1968: 50 f.). and nion, been convincingly argued that thev could not have mixed this may account for some of its so-called Kurgan traits (Gimbu­ in a combined process of migrations and local processes of tas 1979). However, neither Hausler in his recent work (1983). change. On the contrary, C14 dates seem to support a rapid nor others are able to point out a specific region of origin for process of migrations and acculturation from 3000: 2900 B.C. e.g. the Corded Ware potterv (a recent summarv bv Buchvaldek in Eastern Europe to 2800 B.C. in Jutland and Northern Euro- 216 — — Niederlande Dcinemark SCidw. d. BRD Mitte/etbe - Saa/e - Gebie t Obertousitz. £i»ielgrobkultur btw. Schnurkeratпік bfockenbecher- Aunjetitzer Schnur- Kultur Kultur Kerami к +o r 9 ^ r <*L9 D n л г / L L (? cf ^їїЖіяа bsssssi^ ^ d 1EBJESSn ЙВИ!^ 2 d 1““* “““і ^ 9 Mecklenburg HaffkCtsten- MaAopolska Fat'janovo- Schtredische Estnische MitteMnepr- Kulfur Киї Ти г SfreitaxtkulTur Kultur Schnurkeramik bzw. Einzelgrabkulwr (Nordqruppe) 9г? 9г / Т 9 dГ cгf? cгf L L X / ' D L L L L сГ? cf O' 9 9? 9 Fig. 2. Schematic presentation of basic principles of burial positions in the Late Neolithic/Early Bronze Age Cultures in northern Eurasia (after Hausler 1983). pe (Gimbutas 1979, Pape 1981). This is reflected in the so-cal­ only to central and northwestern Europe, but also to Sweden, led “gemeineuropaisches Horizont” (Buchvaldek 1986 for a re­ Finland and eastwards into the USSR (the Middle Dnieper and cent status), which by many scholars is taken to represent an ini­ Fatjanovo groups, e.g. Ozols 1962) that would seem to favour tial migrating phase from the east, originating at the interface an eastern rather than a central or west European source. Since between Kurgan and Late Neolithic Cultures between the most groups of the Corded Ware and Battle Axe cultures are Dniepr/Dnestr and the upper Vistula/Oder7 (Sulimirski 1968: chronologically synchronous (Pape 1981), this also suggests ra­ 84 ff., map VIII and X; Buchvaldek 1985), which in terms of diation in many directions from one or a few regions of origin. topography forms a natural continuum. It lends support to the Anthropology: Although the sample is uneven and small it hypothesis that we also find this phase in Jutland, and that it re­ seems clear that no specific anthropological type is linked to presents a genuine case of tribal migration, actually the conclu­ the CWC in Europe. The material varies regionally and locally, sion of a common-European migration. As has been pointed but is generally dominated by the more gracile Mediterrannean out correctly by Buchvaldek recently, a first generation of sett­ type, supporting an autochthonous tradition. Some “kurganisa- lers are not likely to leave much evidence, compared to succe­ tion” in populations can be observed in eastern Europe, de­ eding generations (Buchvaldek 1985: 488, Abb. 3). Hausler has creasing westward and is not observable in western Europe further demonstrated that in terms of burial positions the SGC (Schwidetzsky 1980). This is by some taken to account for an in­ belong to a northern Eurasian complex characterized by strict flux of “kurgan” populations throughout the neolithic of Euro­ divisions according to kin, sex and age (fig. 2).8 On this pe (the three “waves”: Gimbutas 1979), but should probably be background it seems probable that recent excavations in both restricted to account for the small, well defined “Kurgan” popu­ Jutland and the steppe region may reveal further similarities in lations penetrating into Eastern Europe (e.g. Ecsedy 1979). terms of burial constructions and ritual, e.g. house or hut con­ The Kurgan type is anthropologically characterized as tall structions, if subjected to systematic, comparative analyses. (average of males 173 cm), robust and with curved forehead. Nature of expansion: Also the nature of geographical spread Two variants are distinguished: proto-Nordic and proto-Cro- of the Corded Ware and Battle Axe cultures lends support to Magnon. Again a marked border is the river Theiss, east of this model (fig. 2): in Central Europe small pockets of local which a Kurgan population can be defined, apparently most Corded Ware groups at rivers and in vallev systems, with large clearly among males, whereas the women are often of the “empty regions" between them where resistance was too strong, gracile European/Mediterrannean type, reflecting marriages alternating with regions of more massive expansion where con­ with local women (Marcswik 1979). Also the Baltic group is ditions were favourable or receptive, as in the North European distinctly different from the common European type. lowlands characterized by the Single Grave/Battle Axe Cultu­ In Denmark (and Sweden) a marked increase in mean res. To this mat be added an often observed, but none the less height (7-8 cm) can be observed between the TRB and the significant geographical dimension: the massive spread not Late Neolithic Dagger Period (Bennike 1985: fig. 13). which 217 may have taken place during the Single Grave period. Unfor­ ture in northern and western Europe, later to be followed by tunately most skeletal material originates from Eastern Den­ other climatic fluctuations. It may not then be accidental that mark, that is outside the original Single Grave core region, and the climatic decline around 3000 B.C. in much the same way here continuity prevails. A recent Danish find with preserved saw' an expansion of new social and economic practices. Such a Single Grave skeletons from Jutland, however, seems to suggest climatic change may have favoured grassing of more marginal anthropological traits falling outside the normal Neolithic soils and may have caused problems in the traditional agricul­ range, which may be supported by a reanalysis of some of the tural communities, establishing a pattern of long term oscillia- Swedish evidence (Petersen 1988, During 1989), but we must tions between central and marginal agricultural regions await more complete, up to dage analyses, just as we need more throughout Europe (discussion in Whittle 1982). In Scandina­ material from the SGC. Since very few skeleton remains are via the effect of climatic deterioration in the later 4th millen­ preserved from the SGC we cannot evaluate the impact of new nium is demonstrated in the southward retreat of agriculture populations compared to other factors, e.g. changed or im­ and the concommitant expansion of hunter/fishers of the proved diet. Especially milk products tend to raise the height. Pitted Ware/Comb Ware tradition from northern Eurasia A recent survey of dental conditions of Single Grave/Battle Axe (Graslund 1981). Thus the Pitted Ware and the Battle Axe material in Denmark and Sweden has indicated some differ­ cultures could be suggested as representing different but inter­ ences compared to other neolithic groups. Especially “enamel acting responses to large scale ecological and climatic changes hypoplasia”, which reflect periods in young age of starva- in respectively the forest zone and the steppe zone of Eurasia. don/bad nutridon or periods of illnes and fever, was rare, It should be stressed, however, that no climatic determinism suggesting excellent living conditions for the buried popula­ is to be implied. Social and economic dynamics determine the tions, probably due to a dominance of milk/meat products course of development whereas ecology and climate set bar­ (Alexandersen 1989: 176). riers that are sometimes transcended. While anthropological data from Denmark does not at pres­ Technological and economic innovations. The importance of the ent allow the identification of a migration, due to lack of repre­ secondary products revolution as a contributory factor to the sentative data, it seems clear that there existed larger regional spread of the Corded Ware/Battle Axe Cultures is perhaps differences throughout Europe rooted in a more remote past somewhat overstated. As indicated by Sherrat (1981 table 10.9 (Menk 1980). In the Late Neolithic/Early Bronze Age (Aun- & 10.10) both ox-traction, ard ploughing and probably also jetitz) there is apparently a distinct difference between Scandi­ wheels and milk products belong with the TRB Culture of the navian and Central European populations. Anthropological 4th millennium. What is left for the Corded Ware and Single evidence does then seem to rule out Central Europe as the Grave Cultures is mainly the exploitation of sheep for wool and origin of the Battle Axe/Single Grave Culture in Denmark and perhaps horse riding. This had some social and economic con­ Sweden. sequences: sheep tolerate soils of lower quality and the horse Indigenous hunter/galherers. The question of a surviving meso­ allowed more rapid social interaction. Another basic precondi­ lithic substratum is an old one, but has recendy been revived by tion, however, was the combination of open landscapes in the especially Rrzak (1981) proposing that the Corded Ware com­ settled areas and still large tracts of unsettled secondary soils. plex originated in areas of mesolithic tradition in Europe. This Thus the scene was set for wide scale changes and interactions has been firmly rejected and, it seems, quite righdy (e.g. Haus- between the semi-arid steppe regions of Eurasia and the lightly ler 1983). On the other hand there are a number of traits in forested secondary soils in Europe, as convincingly demon­ burial customs of the Ochre grave/Kurgan cultures that bear strated by Sherrat (1981: 295 ff.). This included both actual resemblances to mesolithic traditions of northern Eurasia, as migrations and the spread of new linguistic and social systems observed by especially Hausler in his works. In the early third (Sherratt and Sherratt 1988).9 It seems to me that Sherratt's in­ millennium there existed two major cultural complexes in this terpretation more satisfactory than others accounts for the region: in the forest region the Combed Ware/Pitted Ware- available evidence and allows a combination of both Shennan's groups of semi-neolithic hunter/fishers (W'yszomirska 1984: and Gimbutas' models. fig. 5) and in the steppe region the Ochre Grave/Corded Ware Although we may be able to define and locate some of the cultures. Quite naturally they interacted throughout their fron­ constituing elements in the development of the CWC, it is not, tiers in the Baltic and in the USSR. It is remarkable, however, at present, possible neither to describe nor explain its precise that thev bodi expanded westwards around 3000 B.C. - the origin, although Sulimirski (1968) and Buchvaldek (1985) Pitted Ware shortly before, the Kurgan/Battle Axe Cultures have made convincing proposals (fig. 3). One reason for this is shortly after. This leads on to the impact of climatic change. already mentioned - the general lack of theoretical and analy­ Climalic change. A number of independent climatic observa­ tical sophistication in Corded Ware studies. Another is the tions suggest that a climatic change towards cooler and wetter neglect of coming to terms with the nature of population climate occurred around 3000 B.C. or shortly after. (Aabv 1976, movements. Neither parties have taken into account the com­ articles in Harding 1982, W'igley et al. 1981, especially Bowden plex nature of migrations. First: л migration need not take place et al. 1981: fig. 21.2 &• 21.3 and Porter 1981). The first signifi­ in one sweep, but can act as a kind of catalyst, one group forcing cant climatic detonation after the Boreal Atlantic optimum oc­ others inbreak up. thereby dissolving the image of a unified ma­ curred around 4000 B.C. coinciding with the advent of agricul­ terial culture. Second: migrations put very specific demands 218 Fig. За. Geographical model for the genesis and expansion of the Corded Ware/Battle Axe cultural complex (after Gimbutas 1986). b. Geographical presentation of Corded Ware/]amna cultural groups, and a proposal for the origin and spread of the Corded Ware/Battle Axe cultural complex. 1) Distri­ bution of Corded Ware Culture Groups; 2) Jamna Culture; 3) presumed area of origin; 4) presumed main directions of the primary distribution; 5) CW in Central Europa; 6) CW in Switzerland; 7) CW in NW Europe ("Single Grave Culture" in Denmark and NW Germany, "Standvoetbeker" culture in the Netherlands, CW on the lower Oder river; 8) CW in Great Poland; 9) "Rzucewo- Baltic Haft Culture"; 10) CW in the West Ukraine and in SE Poland; 11) East Baltic of "Boat-shaped axes"; 12) Middle Dnieper group; 1 3) Fatjanovo and Balanovo groups (after Buchvaldek 1980). 219 upon those involved leading to specific changes and adjust­ Cimbric/Teutonic, and several North American and African ments in material culture and social organisation. Consequent­ historical migrations). This, however, also depends upon the ly we cannot expect a priori to find a unified material culture scale and nature of migrations to which we shall now turn. between areas of supposed origin and areas of final settlement. Contexts and types of population movements. In the following we Thus the selective mechanisms operating under the specific distinguish between full scale and select movements. Other cri­ conditions of migrations have never been systematically analy­ teria may be employed, such as speed and directionability, but zed. We here need comparative studies under historical con­ they belong with a later archaeological discussion about the na­ trol. In order to improve on this situation I shall finally discuss ture of migrations, e.g. differences between settlement expan­ the identification of migrations and after that try to outline sion and population movements and the possibility of some types of migrations and their contexts. It is a heuristic distinguishing between them (Neustupny 1981). sketch without any attempt to cover or refer more than a seg­ 1) . The full scale movement of social groups may be diffe­ ment of the relevant literature. rentiated into three types: - displacement by states/empires - social conflict/tribal competition - ecological/economic pressure A TENTATIVE SCHEME FOR POPULATION MOVEMENTS The full scale movement of tribal groups, including children, We have until now employed the traditional concept “migrati­ livestock etc. is not very well documented in prehistory, except on”. As it bears many simplistic and value loaded associations for Caesar's description of the Helvetii, and some Celtic and Vi­ with it, one might prefer the more neutral expression “popula­ king immigrations. The Single Grave Culture, however, probab­ tion movements” to stress the diversity of such movements - ly belongs here. It should be pointed out that even such large from individuals over select groups of traders and warriors to migrations did not normally deprive a region of its population, whole populations. In the following we are mainly concerned but rather represented the combined effect of several groups with movements of larger groups of people. Although “large” or settlement units joining together, as it is known from the Vi­ remains a relative word, it serves to differentiate between indi­ king period. We should not therefore a priori expect major dis­ vidual movements of marriage partners, mercenaries, traders placements or decline of setdement in those regions providing or settlements from the co-ordinated movements of a group of the people, although it may sometimes occur. people, whether voluntarily or forced, to occupy a new area. It Causes include political displacement of oppositional ethnic marks a not easily definable difference between “interaction” groups, a policy followed by all empires throughout history (the and “take over” (or at least an attempt to take over) by moving Jews in Babylonia, Celtic and Germanic tribes by the Romans in larger groups of people, whether farmers, traders, warriors etc.), internal social conflict/exclusion (part of the Viking ex­ or all at the same time. There are obviously intermediate stages pansion, e.g. Eric the Red and his group leaving for Greenland, of various types of domination. some of the Polynesian expansion), political subordination or The first problem confronting any study of such phenomena the threat of it by intruding dynasties (several Iron Age migra­ is that of identification. tions), social and ecological constraints (the Corded Ware Cul- Identification includes three elements: ture/the Single Grave Culture in Jutland, the Cimbric/Teuto­ - intrusion of an alien group (resettling) nic migration from Jutland), and planned migrations to take up - a migratory route (connection) new land (the Helvetii as described by Caesar, several of the - a mother culture (origin) Pueblo cultures in the American Southwest). Historically the process operates in the opposite order to this, 2) . Select movements of social groups may be divided into at but identification in the archaeological sense is mostly reversed, least four variants: beginning where the process stops; and for very good reasons, - conquest since the replacement of one culture by another is often the - mercenaries most conclusive evidence archaeologist can come up with (e.g. - trading stations/colonies the Single Grave Culture). As previously indicated it may take - labour/stigmatized groups many forms from virgin settlement to various types of cultural Migrations of select social groups is probably as widespread in mix, the processes of which are still badly known. prehistory as in history. It includes the intruding of foreign Migratory routes are the most difficult to trace due to the se­ chieftains/kings and their retinue taking over control - so-cal­ lective mechanisms at work. In several well documented cases led conquest migrations. Examples include the recurrent in­ both the mother culture and the final region of settling down flux of nomadic groups in Europe, from the Scvthians, the can be traced, whereas the route is only represented bv scatter­ Huns to Genghis Khan, some of the Tumulus expansion of the ed finds (e.g. the Bastarnae: Schlette 1977, the Cimbric/Teuto- Middle Bronze Age, part of the the Celtic and Viking expansion nic migrations: Sever 1976 Abb. 51, several Germanic migrati­ and the widespread feature of intruding dynasties in the myths ons: Kriiger 1977. the Langobards: Werner 1962). Thus, it is of origin of African kingdoms. This may either result in a fast clear that without literarv sources a number of well known mi­ acculturation or an influx of larger groups from the home base grations could not have been identified archaeologically, at of the new leaders, which could explain part of the Nordic least not in our present stage of knowledge (e.g. the Bronze Age expansion. Also mercenaries probablv belong 220 here, since on return they often bring with them strong influ­ When dealing with the structural and cultural framework of ences, such as the Germanic mercenaries in the Roman army. migrations we also have to face the problem of ethnicity and From historical sources we have ample evidence of such mi­ language (Barth 1967, Hodder 1982, Herrmann 1988) and litary/political movements and take overs (for comparative dis- their relationship to material culture, since a strong element of cusison, see Webb 1975), that have often left rather weak traces ethnicity is implicitly assumed in the identification of migrati­ in the archaeological records (e.g. the Huns as documented by ons. This includes the difficult problems of “ethnogenesis" Werner 1956, or the Vikings). This is not at all surprising given towards whose prehistoric reality 1 remain sceptical.10 We the nature of such migrations. What has left an impact, how­ should be aware that ethnicity may often be exclusive and class ever, is often place names, since they symbolize the political/ad­ defined and that the modern notion of ethniticy as all-embra­ ministrative take-over of a region. Thus place names are prob­ cing (a people) may not be applied uncritically to prehistory. ably a good indicator of succesful conquest migrations. We should also be aware that different ethnic groups may co­ Another type is represented by trading stations/colonies, exist, as demonstrated by Barth (1967), often as a result of which is often characterized by the same lack of a clear archae­ migrating groups taking up new niches. ological identification, or of a mix, such as the Vikings in Rus­ Thus, our ability to identify and understand ethnicity, sia, the Myceneans/Greeks and Phoenecians in the Meditar- population movements and language in the archaeological re­ rannean (Kimmig 1982). cord depends primarily on our ability to identify and explain Finally there are the eternal migrations of stigmatized ethnic the social and economic framework within which they opera­ groups taking up specific tasks, such as blacksmiths/potters, ted, as recently suggested by Colin Renfrew (1987). But even trade and barter (jews/gypsies), labour (slaves), which is to be that will not do. If we want to see some future advances in the seen as a more permanent structural outcome of large scale study of population movements, it is necessary to carry out processes of ethnic displacement and exploitation in empires comparative studies under historical and contextual control in throughout history. order to specify those conditions under which they may occur From the above observations a certain processual develop­ and the selective mechanisms at work during processes of repla­ ment in the type of migrations may be suggested, tribal migra­ cement and change. This is a precondition for understanding tions of the neolithic mainly caused by ecological/demogra- and identifying the material correlates of various types of popu­ phic problems, to trade and conquest migrations/further mi­ lation movements. grations of conquered peoples in later periods ffom-the Bronze Age onwards. This also implies differential impact upon mate­ Kristian Kristiansen, Ministry of the Environment, the National Forest rial culture. Such a scheme was recently proposed by Colin Ren­ and Nature Agency, Slotsmarken 13, DK-2970 Horsholm. frew, although he preferred to fix the evolutionary fault line for elite domination with the beginning of the Iron Age (Renfrew 1987:131 ff.). We should, however, be cautious not to apply ex­ Acknowledgement cessively simplistic models. Tribal chiefdoms may very well from This paper was first presented at the second Nordic conference on the an early stage have migrated according to a model of elite do­ Battle Axe Period, held in Lund, Sweden, 31st October to 2nd Novem­ mination, just as migrations without any rational demographic ber 1988, and later in England at the TAG conference in Newcastle, De­ or ecological background may be found, e.g. to seek mythical cember 1989, in a session organized by this author titled: “New Perspec­ origins or just sheer explorations of new land. In most cases that tives on Prehistoric Migrations”. The lively and constructive responses we know of, there was already a familiarity with the new lands received on both occasions convinced me that time was ripe to re-open either through exchange, trade, alliances or explorations the discussion about prehistoric migrations - 40 years after the publica­ (Helms 1988). tion of Childe's “Prehistoric Migrations in Europe”. Although I have In conclusion the archaeological identification and the cul­ maintained the paper mosdy unchanged, the contributions from both tural and structural contexts of migrations or population move­ occasions have helped to clarify some of the more crucial questions and ments are still badly understood. They depend on the level of formulations. Finally 1 want to thank David Liversage for correcting the social and political organisation and their interaction with English. demographic, economic/ecological and political factors. Some cases are rather clearly due to ecological/demographic pro­ NOTES blems (some of the Iron Age migrations, e.g. the Cimbric/Teu- tonic migration), but in more advanced stages of social and po­ 1. A note of clarification: diffusion is a covering, descriptive concept litical organisation military conquest may also force subdued for the transmission and change of material culture that does not groups to migrate. Even internal contradictions and competiti­ account for its underlying social mechanisms. Migration is likewise on is known to have led to migrations (e.g. Eric the Red). States a covering concept for the movement of people, from whole popu­ and empires reallocate whole ethnic groups, just as they in pe­ lations to smaller groups. Here. too. the underlving social mecha­ riods of crisis are tempting centres of wealth for migrating nisms are not accounted for. tribes from outside the empire (e.g. the Germanic people and 2. A few indeterminable structures have been excavated. One is circu­ the Roman empire). All this is historically well known, hut has lar and certainly not a house (Rostholm 1986). Another consists of attracted onlv little attention from archaeologists. lines of stake holes (Liversage 1987. lig. I). resembling similar