Gosden 1985 Gifts and kin in the early iron age Europe

20
Wiley and Royal Anthropological Institute of Great Britain and Ireland are collaborating with JSTOR to digitize, preserve and extend access to Man. http://www.jstor.org Wiley Royal Anthropological Institute of Great Britain and Ireland Gifts and Kin in Early Iron Age Europe Author(s): Chris Gosden Source: Man, New Series, Vol. 20, No. 3 (Sep., 1985), pp. 475-493 Published by: Royal Anthropological Institute of Great Britain and Ireland Stable URL: http://www.jstor.org/stable/2802442 Accessed: 07-10-2015 14:49 UTC Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTC All use subject to JSTOR Terms and Conditions

Transcript of Gosden 1985 Gifts and kin in the early iron age Europe

Wiley and Royal Anthropological Institute of Great Britain and Ireland are collaborating with JSTOR to digitize, preserve and extend access to Man.

http://www.jstor.org

WileyRoyal Anthropological Institute of Great Britain and Ireland

Gifts and Kin in Early Iron Age Europe Author(s): Chris Gosden Source: Man, New Series, Vol. 20, No. 3 (Sep., 1985), pp. 475-493Published by: Royal Anthropological Institute of Great Britain and IrelandStable URL: http://www.jstor.org/stable/2802442Accessed: 07-10-2015 14:49 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

GIFTS AND KIN IN EARLY IRON AGE EUROPE

CHRIS GOSDEN

Australian National University

This article considers the social structures of Iron Age Europe north of the Alps and their effects on the production and exchange of items which end up in the archaeological record. The first part of the article presents a review of one explicit model of Iron Age society put forward by Frankenstein and Rowlands (I978), in which trade is seen as a crucial force, and the anthropologi- cal work on which it is based. Historical and linguistic records from Ireland and Britain are surveyed to draw up a different picture of Iron Age socety in which control oflocal production is more important than trade and exchange in defining social position. A brief archaeological case study is presented from northwest Bohemia to illustrate the importance of production, rather than foreign trade, during the early Iron Age (c. 6So-3oo B.C.). Finally, the general implications for the role of trade are brought out and the relations between different areas of Europe are considered.

A main hindrance to understanding the archaeological material from prehistoric Europe is our ignorance of the form of society under study. We know little about the nature, size and scope of the basic units of kinship or how the relations both within and between groups have affected the production and exchange of objects which end up in the archaeological record. This article discusses the possible forms of society which can be glimpsed in historical records and literature from Ireland, the reports of classical writers and the evidence of linguistics. I shall argue that the network of small, independent clans revealed by such sources can be used to make sense of developments during the Iron Age, evident from the archaeological record, such as the increase in the production of luxury items between c. 750 and 400 B.C.

As far as the European Iron Age is concerned the one attempt to present an explicit model of the formation, alliance and ranking of social groups is the prestige goods model put forward by Frankenstein and Rowlands (I978). The prestige goods model looks at the effect that the trade with the expanding states of the Greeks and the Etruscans had on competition between the communities north of the Alps, with especial emphasis on the use to which rich Mediter- ranean items were put in this competition. The principles underlying the prestige goods model were derived from work on such societies as the Kongo of west central Africa and Melanesian groups. Ideas on alliance and exchange similar to those in the prestige goods model were given a more European twist by Rowlands (I980), who derived a putative outline for an Indo-European social system from linguistic analyses and the Homeric literature, before applying this to the archaeological material from the Bronze Age.

The main group to provide inspiration for Frankenstein and Rowlands's Man (N.S.) 20, 475-493

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

476 CHRIS GOSDEN

model of Iron Age Europe was the Kongo ofwest central Africa (although this is never stated by them), which from the end of the fifteenth century came under the influence of Portuguese trade and colonial ambitions. As is discussed below, Ekholm (I972; I977) has shown how those in political power in the Kongo used the system of alliances between groups to channel items obtained from the Portuguese to maintain the structure of the Kongo kingdom. Implicit in Frankenstein and Rowlands's application ofthis model to Iron Age Europe is the idea that a similar alliance structure and political hierarchy operated in both examples. This article hopes to demonstrate that European Iron Age groups did not have the types of marriage alliances found in areas such as west central Africa on which the prestige goods model is based and that, lacking a stable system of alliances, it was more difficult for those politically dominant to support their power through outside trade.

The argument concerning social structure is an abstract one, but it does have demonstrable archaeological consequences. One of the most notable features of the early Iron Age is the upsurge in the production of luxury items from Hallstatt D onwards, possibly under the influence of the Mediterranean. In the final part of this article I present a brief case study from northwest Bohemia, which concerns the adoption of one new production technique, the potter's wheel. The introduction of this technique represents one facet of the increased production of luxury items and illustrates a basic fact about Iron Age societies: in the absence of stable alliances between groups it was the control of production which was the basis of political power, rather than the control of trade. Trade with outside groups, such as those from the Mediterranean, was important only in a few areas and for short periods during which close contacts were main- tained. It is the role of production versus trade which forms the central theme of this article.

The prestigegifts model The prestige gifts model (Frankenstein & Rowlands I978; Friedman & Row- lands I977; Friedman I98 I) and the works upon which it is based (Ekholm I972; I977; Meillassoux I960; Strathern I97I) are well-known and there is no need to give a detailed summary here. I shall therefore only discuss the features which are of particular present interest.

The basis of the prestige gifts model is that power is not based directly upon the regulation and exploitation of the production of food and other basic necessities, but rather rests on the controlled movement of socially important items, the most prized of which are often obtained in external trade with another group organised on different social and economic principles. In the case of the Kongo of west central Africa cloth, guns, beads and mirrors were obtained from the Portuguese after I482 and replaced original prestige gifts deriving from trade with neighbouring states. Prestige gifts were used by the Kongo to maintain a political hierarchy which linked together the basic units of society, which were villages each inhabited by a lineage, to form a state. The different ranks within the political structure, which ranged from king, provincial chief, district chief down to village chief, were marked by access to certain forms of wealth. The

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

CHRIS GOSDEN 477

most important gifts were obtained by the king through long-distance exchange and were passed down through the ranks to maintain the allegiance of the chiefs. The downward flow of gifts was matched by tribute of local specialities gathered by the chiefs and offered ultimately to the king.

Before the Europeans introduced their imports into the area in search of slaves, there had been a considerable and varied local production of politically important items. The primary traditional prestige product was shells, which were gathered and made into necklaces on the coast. The king had absolute control over the production and exchange of shell necklaces and they were passed to him as tribute up the political ladder. Metal tools were also of some significance and were often circulated without being used, indicating that they were worth more as gifts than as instruments ofproduction. Metal working was a privilege of the aristocracy and the first Kongo king was thought to have been a smith. The metalworkers may have been the only full-time specialists in the Kongo kingdom and probably lived and worked in the royal court. As the ore deposits were controlled by the king he had power over every stage of the process from extraction to exchange. Salt deposits and trade were regulated by the king and the otherwise poor province of Soyo had become wealthy due to its salt deposits. The production of the lowest level of prestige item, such as cloth, was not directly controlled and weaving took place in the individual villages, rather than at the court. The more politically important activities of wood and ivory carving took place in the court where they could be overseen by the king (Ekholm I972: II9-26).

The traditional situation amongst the Kongo was of craftwork being carried out in a number of regions in the kingdom in order to produce prestige gifts. The widely distributed production made it difficult for the king to control exchange, a serious drawback to a person whose power was based on the regulation of prestige gift production and exchange. When prestige gifts from Europe started to circulate, the local production of rich items seems to have declined and then stopped within the Kongo kingdom. Iron and copper extraction ceased, a lesser quality and quantity of cloth was woven and salt production declined (Ekholm I972: I27). The fall-off in production may have been because the king found it easier to maintain the monopoly control of outside gifts which entered the kingdom through a small number of routes, than to keep track of a variety of local crafts. The political importance of the monopoly over European imports is demonstrated by a letter the Kongo king wrote to his Portuguese counterpart, urging him in the strongest terms not to have any trade dealings with Kongo chiefs (Ekholm I972: IOI).

Unfortunately for the Kongo king the monopoly over trade, which formed the basis of his power, was directly counter to the Portuguese interests. The Portuguese were engaged in commercial trade and wanted to encourage com- petition amongst their clients in order to ensure the highest prices for their goods. Consequently the monopoly over prestige gifts, which the Kongo political hierarchy relied upon, was eventually broken and the king's power dwindled. As Ekholm herself explains:

If the king's monopoly over external trade is bypassed and his vassals are able to acquire European prestige articles through other channels, which is what happened after the initial phase of contact,

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

478 CHRIS GOSDEN

the established hierarchy breaks down. As prestige articles are transmitted to the local groups from outside and not via the hierarchy, the basis for this hierarchy disappears. We get secessions from the macro-unit as the local groups began competing and fighting with one another. The kingdom of Kongo collapsed in the middle of the i6oo's ... (I977: I3I).

What is stressed here is that the Kongo may have given up local production of rich items because Portuguese items could be circulated in a controlled manner through the alliance network. If this is true then it may indicate that Kongo society was rather differently ordered from Iron Age groups who responded to outside influence by stepping up production.

Prestige gifts and the European Iron Age Frankenstein and Rowlands used the general principles seen to operate amongst groups such as the Kongo to interpret developments in Baden-Wiirttemberg and the Hunsriick-Eifel during Hallstatt D and La Tene A. From the burials in Baden-Wiirttemberg during Hallstatt D, for example, they saw four tiers of chiefly rank-paramount, vassal, sub-chief and minor chief. Each tier was identified from the construction of the graves and the local or exotic artefacts accompanying the burials. The paramount chief had an elaborately constructed grave, which contained a wagon, horse trappings, imported Greek and Etruscan pottery, as well as bronzes, gold, silk, amber and coral, together with rich local items. At the other end of the chiefly scale were the minor chiefs, buried without a wagon or imports, but accompanied by local products such as daggers, belt plaques, fibulae and pottery. The geographical limits of the chiefly domains were discerned from the areas of distribution of certain stylistically distinct local artefacts. The outline of Hallstatt D society in Baden-Wiirttemberg combined a spatial and a social aspect which distinguished the paramount chief, who was associated with the large settlement at the Heuneburg and linked this position to the minor chiefs within their local domains (Frankenstein & Rowlands I978: fig. i, table i).

Frankenstein and Rowlands feel that there were similar patterns of society in Iron Age Europe and the Kongo, without remarking upon the differences between these groups, modes of life. Although certain of the similarities be- tween the two groups are striking, it is the variations in their reactions to outside influence which may provide the main clues as to how they operated. I shall consider two main areas of difference: Iron Age production and the system of kinship and alliances.

Iron Age production The Greek's first main period of contact with Europe north of the Alps began during Hallstatt C (c. 750-600 B.C.). The flow of Greek products increased markedly during Hallstatt D (600-475 B.C.) and this was partly due to the establishment of colonies, notably Massilia around 6oo B. C. Etruscan trade rose in La Tene A and continued until the beginning of La Tene B, supplanting Greek contacts. The main period of imports (650-300 B.C.) coincides with the rapid

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

CHRIS GOSDEN 479

developments in the production of non-utilitarian objects in temperate Europe. Categories of metal items, such as fibulae, were produced in much greater numbers and diversity of forms than ever before. In Hallstatt C only a few hundred fibulae are known from the whole of Europe, whereas there are many hundreds from the Hallstatt D and La Tene A phases of one site such as the Diirrnberg (Pauli I978). A new style of decoration and form characterises La Tene A, helping to demonstrate the new inventiveness of the smiths and potters, who are using newly sophisticated techniques and borrowing inspiration in part from the Mediterranean (Jacobsthal I944). The potter's wheel was first intro- duced in Hallstatt D, initially to produce copies of Italian imported pottery on the large settlements such as the Heuneburg and Chatillon-sur-Glane (Lang I974; Magetti & Galetti I980). By La Tene A wheel-turned pottery had spread to many areas and was found on all classes of site. Wagons, which necessitated the co-operation of a number of different crafts, are found more commonly in graves dating to Hallstatt D and La Tene A than previously and this may indicate that the production of these prestige vehicles had increased.

All these strands of evidence indicate higher levels of production, employing new techniques and more complex patterns of co-operation between crafts. Unfortunately we know almost nothing of the farming systems of continental Europe at this time. It is therefore impossible to gauge whether the upturn in craft production paralleled and arose out of an intensification of farming. It seems unlikely that such a growth in production occurred by coincidence at the same time as increased Mediterranean contact; a causal link between the two appears more likely. It has been argued that prestige gift economies are, or were, found in many parts of the world (west Africa-Ekholm I972; I977; Micronesia, Polynesia and Melanesia-Friedman I98 i). Frankenstein and Rowlands contend that the Iron Age groups provide one more specific example of the general model. However, the fact that production was stepped up around the time Mediterranean influence was first felt indicates that Iron Age societies may have differed in important ways from other societies grouped within the general prestige gifts model, which reacted differently as far as production was concerned.

Social structure in Iron Age Europe There is evidence that group structure and marriage alliances in later prehistoric Europe were ordered rather differently from the practice amongst the Kongo. In order to consider the kinship structure of Iron Age Europe in more detail I will draw from four different sources: the early Irish laws, early Irish literature, the writings of classical authors and linguistic analyses. The main source of rel- evance is the early Irish laws; it is only here that any detail can be found on the nature and workings of kingship groups and alliances. My intention is not to say that continental Europe of Hallstatt D and La Tene A was organised on exactly the same principles as Irish society of a much later date; it almost certainly was not. What I should like to indicate is that the laws, when combined with other evidence, are suggestive of a state of society rather different from that specified

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

480 CHRIS GOSDEN

by the original prestige gifts model and that such a social form may have existed in its essentials within prehistoric Europe.

The Irish laws were written down in the seventh and eighth centuries A.D. and consist of the laws themselves, plus glosses and commentaries added in the Medieval period. The first written records were derived from customary laws, which had been preserved by an oral tradition for a considerable period. The complete corpus of legal tracts have not all been reliably translated and edited, although modern translations and editions are regularly appearing (Hughes I972 contains the main sources). As in the prestige gifts model, the Irish society displayed in the laws had a political system which was constructed in the idiom of kinship. The basic political unit was the tuath and it is estimated that in the seventh century A. D. there were at least i So tuatha and a total Irish population of less than half a million (Byrne I967: 45). Although the word tuath means 'a people' one such territory may have contained many different kin groups. The important thing for the position of the group as a whole was the kin affiliation of the king (ri). The king of a tuath was the lowest level of king and was bound to an overking riuri, who had at least two other tuatha subordinated to him, as well as his own. If the king of a tuath could claim common descent with his overking this gave the tuath as a whole greater status and meant that it did not have to pay tribute. However all sub-kings, whether they paid tribute or not, had to acknowledge their inferior status by receiving gifts (tuarastal) from the over- king. The highest king of all mentioned in the laws is the ruler of a province ( ri ruirech). The status of the higher kings depended on the number of tuatha under their control and the degree of subordination they could impose (Niocaill 1972).

Each tuath was made up of a large number of kin groups which were often closely related, but not necessarily so. Marriages usually took place within the tuath; marriages to members of other tuatha were often less prestigious and secure. Children belonged to their father's kin group (except where the father was a 'foreigner'). Women became members of their husband's kin group, but did not lose all their ties to the group of their birth. The continuance of natal ties is shown by the fact that if a woman committed a crime, part of the fine was paid for by her sons and part by her own kin. Feuding also involved both kin groups of birth and marriage (Hughes I972: 46-8).

The basic unit of kin was the derbfine, a group claiming descent through the male line from a common great-grandfather. 'Derbfine' means the 'certain kin' and was distinguished from wider kin groups, such as the iarfine (a five generation group) and the indfine (a six generation group). These larger groups only gained power in cases where the debfine ceased to exist. The derbfine held common arable and pasture and farmed co-operatively. It was the group within which the individual's rights, duties and status were defined and defended.

Charles-Edwards (I972) has argued that the derbfine was the central point at which status, landownership and legal responsibilities met to determine indi- vidual and group standing. He has traced similarities in Irish and Anglo-Saxon societies back to a common pre-Roman Indo-European heritage, from which derived basic and ancient social principles. The most important of these principles was that defining the hide. Charles-Edwards sees the hide as the crucial social and legal qualification, because it was the minimum land-holding

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

CHRIS GOSDEN 48I

needed to ensure free status and full legal rights. The exact size of the hide was uncertain, but it was at least the amount of land which could be worked by one plough team in one day. It was the minimum holding needed to ensure the status of boiare in Irish society and coerl in Anglo-Saxon England. The amount of land held separated a free derbfine from a dependent one. The derbfine land was divided equally between its male members and it was necessary for each man to possess one hide in order for that man and the group as a whole to be free.

In the later law tracts the derbfine was replaced by thegelfine, a three generation unit which would ideally have five male members. The gelfine should have possessed five hides in order to escape servile status. In this period Charles- Edwards saw a series of rules linking land, kinship and status.

A hide is a sufficient and necessary qualification for the status of normal freemen. Possession of a hide distinguishes a freeman from a semi-freeman. Possession of a five hide unit distinguishes a free lineage from a semi-free lineage . .. From another tract we learn that the preferred type of vassalage is one between two kinsmen. The status of a lord requires that the holders of a five hide unit, namely a lineage, should be his vassals. Lineage, lordship, status and land are bound together in a tightly knit unity (Charles-Edwards I972: 2I).

Irish society had a definite hierarchy, at the top of which were kings, followed by various grades of nobles and ritual leaders, under whom were free-farmers and superior craft specialists. One rung lower were those with impaired legal status: the tenant farmers, labourers and lesser craft workers. Lowest of all were the slaves, who had no legal rights and could be alienated along with land. Although gift giving and tribute played a part in upholding the political hierarchy, there were more important bases for social position, the chief of which was the ability to be self-supporting and politically independent. Inde- pendence derived from the possession of means of production of all the necessities of everyday life and could not be easily compromised by the need to render tribute to and receive gifts from those who were socially superior.

Irish society had a different foundation from that put forward for prestige gift groups, such as the Kongo. Among prestige gift groups hierarchy was based on the movement of people and things. Such movements were also important in Ireland, but self-sufficiency, property ownership and maintaining the support of the immediate kin group received greater emphasis. Local production bred independence and limited the influence of the political hierarchy. Charles- Edwards has argued that such a condition of society may derive from a common Indo-European heritage and may long pre-date the first historical records of its existence. That is to say similar societies may have existed in Europe during the Iron Age.

Other indications that the type of Irish society known from the laws had some antiquity are contained within literary works such as the Ulster Cycle. The most important of these is the Tain. The Tain was written down in the eighth century A.D., but derived from a far older oral tradition. Jackson (I964: 44), whilst calling the Tain 'a window on the Iron Age', has argued that it describes a period between the birth of Christ and A.D. 300. Hughes, on the other hand, feels that the Tain has an earlier origin and she notes similarities between La Tene archaeological finds from Ireland and items described in the tale. The ideas of

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

482 CHRIS GOSDEN

status and honour, which provide the basis for the laws, are of paramount importance to the action in the Tain and find echoes in Icelandic and Anglo- Saxon literature. As Hughes notes 'The Cattle Raid of Cooley [the Tain] shows us institutions which we are familiar with in the historic period. . .' (I972: I75), but adds that these are contained within an archaic setting of political groupings totally different from those of the historical period.

Jackson (I964: 43) has noted how closely the society shown in Irish literature is paralleled by the accounts of classical writers of Iron Age Europe. Tierney (I960) recognised that most of the accounts of the Iron Age contained in classical writings were based on Posidonius, whose original works have been lost. The influence of Posidonius is felt even on writers with first hand experience of Iron Age groups, such as Caesar and Tacitus. Caesar gives a simplified account of Celtic social structure with three free ranks of nobles, druids and farmers, together with the slaves. He also demonstrates that there were kings with different degrees of power, so that leaders such as Cassivelaunus in Britain and Vercingetorix in Gaul were able to become temporary war leaders of lesser kings. Both Caesar and Tacitus remark upon the role of cientship in war, where the nobles in their chariots were defended and supported by their client group fighting on foot. Interestingly, Tacitus comments on the similarity of British and Irish society and customs in the time of Agricola, although he almost certainly never had any first-hand knowledge of Ireland.

The final area of research which has contributed some insight into the nature of society in prehistoric Europe is the linguistic analysis of kinship terms. The conclusions drawn from such analyses have a certain consistency. Most workers agree that there were few indications of cross-cousin marriage existing in later prehistoric Europe, but instead prefer to see an Omaha system, which later developed into bilateral kin groups. Friedrich (I966) and Barlau (I976) have both reached this conclusion in their outlines of Indo-European kinship. Goody (I983: appendix i) feels that the kinship groups of ancient Europe, such as the genos of Greece, the 'tribes' of Israel and the agnatic groups of Rome always had bilateral tendencies, but that these were contained within a patrilineal, or sometimes a matrilineal, structure. A basic bilateral structure is also felt to have existed amongst the clans in Ireland and Scotland, which traced descent through the male line, but which acknowledged responsibilities to maternal kin in settling feuds and paying compensations. Charles-Edwards (I97I: I22; I972: I7) has shown that kinship groups reckoning descent through the male line, but which do not display any signs of cross-cousin marriage existed in the Celtic areas of Brittany, Wales and Ireland until these were broken down by the growth of larger political units in the early Middle Ages.

The groupings of ancient Greece have been the subject of an analysis by Rowlands (I980) and he discerned the existence of patrilineal groups with prohibitions of an Omaha type in his appraisal of Homer. Similar conclusions have been reached by Humphreys (I978). Rowlands goes on to view the Bronze Age groupings as being ordered by Omaha principles and looks at the implica- tions of this for the long-distance trade networks of the period. It may well have been that the patrilineal clan with an Omaha marriage prohibition was a social form of wide occurrence in prehistoric Europe and which survived in areas such

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

CHRIS GOSDEN 483

as Ireland longer than in regions influenced by new social and economic institutions. There may have been such social changes by the late Iron Age that the clan structure, which had been a general phenomenon throughout Europe, was only preserved in geographically peripheral areas into the historical period.

The existence of an Omaha system of marriage rules in Iron Age Europe would have definite implications for the operation of social groups. Levi-Strauss referred to this system of marriage rules as follows.

... the generalised definition of a Crow-Omaha system may best be formulated by saying that whenever a descent line is picked up to provide a mate, all individuals belonging to that line are excluded from the range of potential mates for the first lineage, during a period covering several generations. Since this process repeats itself with each marriage, the system is kept in a state of permanent turbulence which is quite the reverse of that regularity of functioning and periodicity of returns which conform with the ideal model of an asymmetric marriage system (Levi-Strauss I966: I9).

He goes on to refer to the Crow-Omaha systems as intermediate between his elementary forms of marriage where alliances dominate and complex forms where multiple links between families and groups occur. The complicated network of links is very different from that put forward for the Kongo, as described by Ekholm.

The Kongo social structure According to Ekholm the local Kongo groups were matrilineal and practised avunculocal residence-'the local segment consists . . . of mothers' brothers and sisters' sons i.e. a group of brothers in the oldest generation plus adult sisters' sons in the following generation. In Kongo, sisters remained members of their own descent group throughout life, but upon marriage moved to their husband's village where their children resided until the day they left their father to take up residence with their mother's brother' (Ekholm I977: i i6).

The localised matrilines formed the kinship organised sub-units bound together by the political hierarchy, discussed above. In terms of marriage alliance the matrilines were related by matrilateral cross-cousin marriage, with women moving up to higher ranking matrilines and men moving down to lower ones (fig. i). Oroups which were wife-takers ranked higher than wife- givers and prestige gifts moved in the opposite direction to wives. The Kongo kingdom was made up of a hierarchy of groups bound together by reciprocal movements of people and things: local produce was passed up as tribute and prestige items were given to subordinate groups through redistribution or marriage gifts (fig. i). Each superior group had several subordinates and must give more than they received-'The highest ranked groups must produce more than their subordinates, and herein lies the importance of the control of natural deposits as sources of prestige goods. The centres of power in Central Africa are located in the vicinity of natural sources of prestige items, copper, salt and zimbu (shells)' (Ekholm I977: I29).

As we have seen above, however, local production in the Kongo system was

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

484 CHRIS GOSDEN

Lineage a

A4;

Lnegesbc

Lineages e,f,g

Prestige goods - Tribute

O O~~Matrimonial alliance

FIGURE I. The alliance system and exchanges of people and things (from Friedman

far less important than external trade in maintaining the pattern of alliances and hierarchy.

But these kingdoms expanded relatively rapidly and crystallised into far more hierarchical polities than warranted by the actual level of surplus production. Even if the central power in Kongo could produce more than the provincial centres (the source material describes them as being about equal in output), the difference cannot be as great as six to one, and it is here that external trade becomes instrumental. Through its exchange relations with X [Portugal], A [Kongo] can change internal products for foreign products and can thus appear to be economically stronger than it is in reality, i.e. in the production sphere (Ekholm I977: I29-30).

The essential points emerging from this discussion are that the local Kongo groups (mothers' brothers and sisters' sons) were linked through a stable and hierarchical system of marriage alliances, in which the preferred partnership was a matrilateral cross-cousin. The hierarchy was maintained by women moving up in marriage to higher ranked groups, whilst men and prestige gifts moved down. The higher ranked groups held their position through access to prestige items; such access was often to foreign trade items which were of greater importance than local items of trade. Group hierarchy therefore rested on access to external trade rather than superior productive power.

Trade andproduction-an archaeological example In the model I have drawn up for Europe political power had a different basis from that put forward by Frankenstein and Rowlands (I978). Compared with

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

CHRIS GOSDEN 485

the kings and chiefs of the Kongo those socially dominant in Europe were disadvantaged on two counts. First, it is likely that fewer foreign gifts flowed into transalpine Europe from the Mediterranean than reached west Africa from Portugal (although comparative levels of trade are difficult to estimate). Sec- ondly, there was no stable system of marriage alliances in Europe through which regular flows of gifts could be maintained to support unequal social positions. The dense set of connexions which existed between groups as a result of marriages meant a complicated and unstable network of 'roads' through which gifts and tribute could flow and these did not form a basis for a stable hierarchy. This is not to say that social differences did not exist in Iron Age Europe; the political structure, however, was based on the control and own- ership of land and the means of production rather than trade; that is on local factors rather than outside influence.

The lack of stable alliances and the importance of local self-sufficiency recorded in the Irish histories go hand-in-hand. The dense set of intercon- nexions between groups was inimical to the use of political power and protec- tive of local group interests. It is noticeable that the political groupings of the late Iron Age and the post-Roman period in Europe, in which the first recognisable kingdoms are formed, were based upon clients and that cientship often cut across kinship lines (Charles-Edwards 1972). It was only by breaking down the kinship groups through cientage that real political hierarchies could be formed. In the early Iron Age the only areas which display social forms similar to those of the Kongo are those which were the main beneficiaries of foreign trade, settlements such as the Heuneburg during Hallstatt D and the resource rich areas such as the Hunsriick-Eifel. Here those with social dominance may have been able to use 'the conversion possibilities of external trade' (to quote Ekholm I977: I30) to overcome temporary constraints that the social structure placed on hierarchy. The kingdom-like forms found around the Hallstatt D centres and the resource rich areas of La Tene A were localised and unstable, as shown by their short duration.

In other areas the role of foreign trade must be reassessed. Although rich items obtained through the long-distance trade networks must have been important in defining local social standing they would not have had power to shape society as a whole, as there were no local kinship structures which could turn advantages in the access to trade gifts into regular social advantages. Ekholm (I 977) has insisted that the Kongo system was living beyond its means, as these were defined in terms of local production, and was only sustained due to the role of foreign trade. If foreign trade can be ruled out as a major determinant of Iron Age social forms, then it is necessary to look again at the part played by production. I shall illustrate the importance of considering production by reference to one area-northwest Bohemia-and to one form of production -pottery manufacture.

From previous studies of pottery production it is possible to draw up some broad parameters as to the type of pottery production undertaken. The level of production will have implications for the nature of demand and the degree of specialisation in production, which in turn can tell us a considerable amount about the type of society under study. Peacock (I982) and Van der Leeuw (I976)

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

486 CHRIS GOSDEN

have both drawn up similar schemes of types of pottery production. They differentiate five broad levels of production. The first of these is household production, where each household makes their own pottery employing very simple technology and techniques of clay preparation, vessel formation and firing. The second level is the household industry, in which pottery is produced for sale or exchange, but on a very small scale and not utilising sophisticated technology. A third possibility is the workshop industry, again producing for sale or exchange and although the potters may not pursue this craft for the whole year they do derive a significant proportion of their livelihood from potting. The workshop industry could be expected to employ a higher level of tech- nology, such as kick-wheels and updraught kilns. The most high-powered forms of production are the manufactory and the industry both of which employ a number of people (a large number in the case of the industries of the modern world) and put a great stress an efficiency of production and cutting costs. Ceramic industries are probably only found in the modern world (Wedgewood's being a well-known example) and there is no evidence for manufactories in trans-Alpine Europe previous to the late Iron Age, when new levels of production and new political forms make an appearance together.

During the period of interest here (Hallstatt D and La Tene A) a new means of pottery production is found in central Europe: the potter's wheel. Local wheel-turned wares are found for the first time on the large centres such as the Heuneburg during Hallstatt D and it appears that they were manufactured on these large settlements under some form of central control (Magetti & Galetti I980). By La Tene A the use of the potter's wheel is found in many areas of

SUDETEN RANGE

1~~OO 000 ~ 00

I lb

FIGURE 2. Physical map of Bohemia. Square indicates area discussed in the text.

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

CHRIS GOSDEN 487

continental Europe, including northwest Bohemia. In Peacock's and Van der Leeuw's terms we may be witnessing the changeover from a household industry to a workshop industry, with more sophisticated technology (the wheel) employed to produce at least part of the ceramic assemblage and with greater specialisation on the part of the potter.

' v

KEY LaTene A-southern import

Hallstatt- La Tene settlement C LaTene A-wagon /horse gear

Surface iron ore V Probably Hallstatt D wagon / horse gear

Evidence of gold working V Probably HallstattC-wagon/ horse gear

FIGuRE 3. The archaeological evidence in Bohemia Hallstatt C to La Tene A.

Turning to northwest Bohemia, the area of interest here, there are other changes visible during this period (more details are given in Gosden I983). Bohemia consists of a shallow basin surrounded on all sides by mountains (fig. 2). Northwest Bohemia (area i on fig. 2) lies to the north ofthe Ohre river and is made up of two parts. The first of these is the Ore mountains, which form the present day border with East Germany and Poland and rise to over I,200 m. There are very few archaeological sites in the mountains. The majority of

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

488 CHRIS GOSDEN

prehistoric settlement lay in a series of basins at the southern foot of the mountains. The western Cheb and Sokolov basins had sandy soil, were of limited farming potential and contained only sparse prehistoric settlement. To the east the Most basin was filled with rich soil and dense prehistoric settlement (fig. 2). This is the main area which is considered here.

FIGURE 4. A stamped decorated wheel-turned pot from northwest Bohemia.

The earliest Iron Age evidence from the northwest is the rich Bylany graves, dating to Hallstatt C (fig. 3). These graves were under mounds in wooden chambers and were differentially furnished, the richest having carts and horse gear, while others had only weapons and pottery in large amounts (Pleiner I978: Obr. I42, I45, mapa 7). Simpler burials were contained in urns with few or no grave goods. There is almost no settlement evidence from this period as yet. In the subsequent Hallstatt D period settlement evidence is abundant (fig. 3). There are many small, undefended farming settlements known which are located near arable land and water. Tumulus burials disappear and in Hallstatt D and La Tene A there are only two possible graves known from north of the Ohre. On the fringe of the area, just to the south of the Ohre, are found rich burials with Greek and Etruscan items of Hallstatt D or La Tene A date. These are the most northerly graves with foreign items known from Bohemia.

The settlement pattern in La Tene A is very similar to that in Hallstatt D, with most settlements continuing from the previous period. The settlement density

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

CHRIS GOSDEN 489

from these two periods appears to be greater than the preceding and succeeding periods (Waldhauser I976). There was a particular pattern of sites during Hallstatt D and La Tene A, with a number of hamlet sized settlements, some of which had single farmsteads as satellites. The hamlets were occupied for a considerable length of time, in some cases from Hallstatt D until La Tene D, but the individual satellite farms were occupied for perhaps a generation only (Waldhauser I977). The larger sites have evidence of craft activities such as iron working, which are not found on the smaller sites (Waldhauser I977: I45-6). The individual farms may have depended upon the larger sites for utilitarian items and this indicates that local exchange may have been of some importance.

Long-distance trade is indicated by the occurrence of Attic black figure ware at Kadan, and painted sherds from six sites in the region (Waldhauser I975). It is not certain whether the painted sherds originated in the Hallstatt area of Austria or from Italy. In either case they demonstrate that northwest Bohemia had long-distance trading contacts in Hallstatt D and La Tene A, despite the lack of Mediterranean bronzes, such as are found in other areas of Bohemia. The occurrence of iron, copper and possibly tin in the Ore mountains may have provided an enticement to foreign traders and it is possible that metals moved south, in the opposite direction to exotic pottery.

The pottery analyses The pottery study was a historical one, which looked at the period prior to the introduction of the potter's wheel as well as the inception of that technique. The potter's wheel was introduced into Bohemia during La Tene A. It was mainly used to make to make bowl forms which often had richly stamped decorations (an example of such a bowl is given in fig. 4). One hundred and ten thin- sections were prepared of Hallstatt D and La Tene A pottery from I4 sites within northwest Bohemia (fig. 5). The mineral fragments contained in the clay were identified under a petrological microscope and the pottery samples were divided into a number of groups on the basis of the minerals contained within them.

During both Hallstatt D and La Tene A coarse pottery was made from materials local to the sites at which the pots were used and the production of coarse wares was unaffected by the introduction of the potter's wheel to produce fine pottery. In Hallstatt D fine pottery (carinated handmade bowls covered in graphite decoration) were made in one centre only (fig. 5, 4), whereas in La Tene A with the introduction of the new technique production had spread to three centres (fig. 5, 4; 2 and i i) and each of the types produced were mineralogically and stylistically distinct. Pots made near sites 2 and 4 on fig. 5 had the widest distribution; those made in the vicinity of site i i had a very localised distribu- tion.

The analysis of the tempers and clays used as raw materials show that the introduction of the potter's wheel meant that more labour was expended on pottery production than previously. The mineral fraction in the La Tene A pots was finer than in their Hallstatt D counterparts, implying that finer clay sources

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

490 CHRIS GOSDEN

were sought out or that more time was spent on clay preparation. In addition new equipment would have had to be prepared and the new technique of wheel-turning mastered. The use of the wheel may be labour saving when it is used for mass-production, but here there is no evidence for an increase of production with the introduction of a new technique. At most sites there are similar numbers of coarse and fine wares on sites of Hallstatt D and La Tene A. Wheel-turned pottery may indicate a greater investment in labour in producing luxury items-not only was production extended to more centres but more time and effort were needed to produce each pot. The advantage was a superior product, made in response to local demands.

7

0

A13 0 5 10 Km

group 1 *group 2

o group 3

FIGURE 5. The pottery groups of wheel-turned pottery arrived at by thin-section analysis.

Sites numbered are- I. Radovesice. 2_ Radovesice Va'penka. 3. Bilina. 4. 2;elenice. 5. Hostomice. 6. 2elenky. 7. Vs'echlapy. 8. Libkovice. 9. Ho'rany IO. Chotovenka. I I. Dob'ric'any. I 2_ Poc'erady I 3. Podbo'rany I 4. Nechranice.

Not only was more labour invested in pots, but the greater time involved in production must have meant that the potters obtained more of their livelihood from potting. There is, however, no indication that pottery production was centrally controlled (unlike the case in the large settlements of Hallstatt D) or that certain areas received more of the product than others. Although more of the labour power of the community was going into pottery production than previously, this was expended at the local level and presumably for local benefit.

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

CHRIS GOSDEN 49I

Demand was from a relatively small area around the points of production, as pottery was not moving more than 30-40 km from the workshop. It is possible to envisage pots moving within kinship groups and in response to kinship obligations. Coarse pottery also reflects local self-sufficiency and any move- ment of the coarse wares was even more limited than that of finer vessels.

Pottery is only one indicator, albeit one that is sensitive to local develop- ments, of changes between Hallstatt D and La Tene A. As noted above, these two periods see a greater density of settlement than in previous or subsequent times. Exchange of metal products between the hamlet sized settlements and their satellites is to be inferred. An increase in the intensity and extent of land use is likely as this would be necessary to support the greater numbers of people and increases in other branches of production such as potting, although no direct evidence of agricultural systems exists at present. The picture that emerges is one of increased emphasis on local production which would be consonant with the picture drawn from the Irish literature in which the maintenance of local social standing and rights depended on self-sufficiency and self-provision. There is little evidence of long-distance trade with northwest Bohemia; the only indications are the potsherds of southern origin. The archaeological record of northwest Bohemia is mainly that of local developments. It could be argued that the increases in settlement density and production within northwest Bohemia were indirect responses to the overall influence of the Mediterranean on central Europe. This would be difficult to demonstrate on the archaeological evidence available and even if true it would ignore the importance of local changes. Any world system, or larger region, only exists as a series of local units in which change occurs and in each area change must ultimately be paid for in the form of production.

Conclusions Trade has (jokingly) been called 'the King Kong of all prime-movers' (Flannery I976: 283). Such a view underlies the prestige gifts model and has received little explicit recognition. Undeniably, trade has the power to shape social formations under certain conditions, as the analysis of the Kongo shows. However, trade is not always socially formative and care must be taken to divide situations in which trade will be important, and even decisive, from those in which it will not. Societies with Crow-Omaha marriage rules may occupy an intermediate state between social forms which have regular alliances channelling the move- ment of people and things, and more complex forms in which exchange is not so influenced by kinship, but has a more commercial nature. In both alliance and complex systems of marriage trade and exchange may have been a prime- mover: in the former case because of the 'roads' created by regular marriage links, in the latter because trade was freed from the constraints of kinship and had taken on a commercial nature.

The Crow-Omaha systems, in the early Iron Age at least, do not seem to have had recourse to commercial transactions, but had scrambled their inter-group links through a complicated system of marriages. They were therefore resistant

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

492 CHRIS GOSDEN

to the build up of regularised trading links, relying more on local production. Later prehistoric Europe may consequently provide the raw materials to study the shift from alliance systems to Crow-Omaha forms (whenever this may have occurred) and finally to complex marriage and commercial exchange, a move which may have taken place with the emergence of the oppida and new levels of production in the late Iron Age. These shifts should be examined in terms of structures of product-ion, kinship links and the effects of trade and exchange.

On a more specific level, the idea that early Iron Age societies were resistant to the effects of external trade may lead to a re-interpretation of links between the Mediterranean and trans-Alpine Europe. The relationship between the two areas is often conceived of in terms of Wallerstein's world system model (Wallerstein I974), where the Mediterranean is the core and the rest of Europe the periphery. Changes within Europe as a whole are seen as due to Mediterra- nean ambitions and influence. Trans-alpine Europe's dependence is basic not only to interpretations of the early Iron Age, but also the later Iron Age and early Medieval periods (Haselgrove I982; Hedeager I977; Parker Pearson I984). If, however, areas such as northwest Bohemia can be seen developing without deep influence from the Mediterranean, perhaps the model should be reversed, or at least turned sideways. The Greeks and the Etruscans may only have become interested in the lands to the north once sufficient productive power had been generated there to make trade worthwhile. Rather than seeing most of Europe jumping at the behest of a structurally dominant Mediterranean, a more balanced view might be that of a series of mutual influence flowing through all areas of Europe. Indeed, if some form of commercial exchange existed in the emerging southern states, this might have made them more susceptible to the influence of outside trade than was the north.

The balance between regions and between production and trade must at least be seen as problematic. Production must have analytical priority over other social influences because all social change must ultimately be supported by people and things, both of which have to be created. At certain times societies may live beyond their local productive means, but the conditions must be right. The identification of these conditions will aid the building of more sensitive and socially complete models.

NOTE

I am grateful to the British Academy for support while this article was prepared. I should like to thank Jim Allen, John Burton, Clive Gamble and Nicholas Peterson for critical comments on various aspects of this article. I am not sure that any of them would want to be associated with the final product.

REFERENCES

Barlau, S. I976. An outline of Germanic kinship.J. Indo-Europ. Stud. 4, 97-I30. Byrne, F. J. I967. Early Irish society (Ist-gth centuries). In The course of Irish history (eds) T. W.

Moody & F. X. Martin. Cork: Mercier Press. Charles-Edwards, T. M. I97I. Some Celtic kinship terms. Bull. Brd Celt. Stud. 24, I-I22.

I972. Kinship, status and the origin of the hide. Past & Present 56, 3-3 3.

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions

CHRIS GOSDEN 493

Ekholm, K. I972. Power and prestige: the rise andfall of the Kongo kingdom. Uppsala: Skriv Service. I977. External exchange and the transformation of central African social systems. In The

evolution of social systems (eds)J. Friedman & M. Rowlands. London: Duckworth. Flannery, K. V. (ed.). I976. The early Mesoamerican village. New York: Academic Press. Frankenstein, S. & M. J. Rowlands I978. Early Iron Age society in southwest Germany. Inst.

Archaeol. Bull. 15, 73-II2.

Friedman, J. I98I. Notes on structure and history in Oceania. Folk 23, 275-95. & M. J. Rowlands I977. Notes towards an epigenetic model of the evolution of'civilisation'.

In The evolution of social systems (eds) J. Friedman & M. Rowlands. London: Duckworth. Friedrich, P. I966. Proto-Indo-European kinship. Ethnology 5, I-36. Goody, J. I983. The development of thefamily and marriage in Europe. Cambridge: Univ. Press. Gosden, C. H. I983. Iron Age pottery trade in central Europe. Thesis, Univ. of Sheffield. Haselgrove, C. I982. Wealth, prestige and power: the dynamics of late iron age political centralis-

ation in south-east England. In Ranking, resource and exchange (eds) C. Renfrew & S. Shennan. Cambridge: Univ. Press.

Hedeager, L. I977. Processes towards state formation in early iron age Denmark. In New directions in Scandinavian archaeology. Cophenhagen: Nationalmuseet.

Hughes, K. I972. Early Christian Ireland: introduction to the sources. London: The Sources of History Ltd, with Hodder & Stoughton.

Humphreys, S. I978. Anthropology and the Greeks. London: Routledge & Kegan Paul. Jackson, K. N. I964. The oldest Irish tradition: a window on the Iron Age. Cambridge: Univ. Press. Jacobsthal, P. I944. Early Celtic art. Oxford: Clarendon Press. Lang, A. I974. Die geriefte Drehschiebenware der Heuneburg 1950-1970 und verwandte Gruppen

(Heuneburgstud. 3. R6m.-germ. Forsch. 34). Berlin: Walter de Gruyter. Levi-Strauss, C. The future of kinship studies. Proc. R. anthrop. Inst. 1965, I3-22.

Magetti, M. & G. Galletti I980. Composition of fine ceramics from Chatillon-sur-Glane (Kt. Fribourg, Switzerland), and the Heuneburg (Kr. Sigmaringen, west Germany. J. archaeol. Sci. 7, 87-9I.

Meillassoux, C. I960. Essai d'interpretation du phenomene economique dans les societes tradi- tionelles d'auto-subsistence. Cah. Etud. afr. I:4, 38-67.

Niocaill, G. M. I972. Ireland before the Vikings. Dublin: Gill & Macmillan. Parker Pearson, M. I984. Social change, ideology and the archaeological record. In Marxist

perspectives in archaeology (ed.) M. Spriggs. Cambridge: Univ. Press. Pauli, L. I978. Der Durnberg bei Hallein III: Auswertung der Grabfunde (Muinch. Beitr. Vor-

Friihgesch. i8). Peacock, D. P. S. I982. Pottery in the Roman world: an ethnoarchaeological approach. London:

Longman. Pleiner, R. (ed.) I978. Praveke dejiny lech. Prague: Academia. Rowlands, M. J. I980. Kinship, alliance and exchange in the European Bronze Age. In Settlement and

society in the British later Bronze Age (eds) J. Barrett & R. Bradley (B.A.R. 83). Oxford: British Archaeological Report.

Strathern, A. I97I. The rope of moka. Cambridge: Univ. Press. Tierney, J.J. I960. The celtic ethnography of Posidonius. Proc. R. Irish Acad. 6o, I89-275. Van der Leeuw, S. E. I976. Studies in the technology of ancientpottery (2 vols). Amsterdam: Univ. of

Amsterdam. Waldhauser, J. I975. Keramiche importy v pozdne halstatskych a casne lateneske objektech

severozapidnich cech. In Rimske Importy. Prague: Academia. I977. Keltske sidliste u Radovesic v severozapadnich cechach. Archeol. Rozhled. 29,

I44-77.

Wallerstein, I. M. I974. The modern world system (2 vols). New York: Academic Press.

This content downloaded from 129.215.17.190 on Wed, 07 Oct 2015 14:49:07 UTCAll use subject to JSTOR Terms and Conditions