From chiefdom to State organization in Celtic Europe

The Bronze Age – age of ambiguity

The principal characteristics of the Bronze Age appear during the middle of the third millennium BC. The Beaker phenomenon is the vector, or one of the expressions, of quite important changes. These include the generalized use of copper across the Continent and the practice of constructing a circular tumulus for an individual burial, sometimes accompanied by a rich assemblage in copper, gold, and amber. Social change is manifested, above all, by the individualization of political power. No longer is it segments of society, such as lineages, that are demonstrating their status through funerary monuments, but rather individuals and their immediate relations. From the beginning of the Bronze Age, certain children’s tombs are furnished with symbols of high social status. Status is therefore probably transmitted through heredity. The dominant social stratum also seems to have preferential access to copper objects. The available historical sources, the spatial distribution of social groups, and their anthropological correlates suggest that they were, for the most part, integrated at an intermediary scale between the local (the site and its immediate catchment area) and the regional levels. Each autonomous political community consisted of around a hundred people on average, distributed in five to eight small settlements. Hereditary transmission of leadership roles is the principal argument for classifying these forms of social organization as chiefdoms. The other argument – socio-economic centralization, which implies the redistributive role of the chief³ – remains difficult to use outside of Wessex as the scale of integration is generally barely above the local level.

This is important because, in effect, it signifies the first stage in the emergence of political organization operating beyond the descent system. If kinship organization offers lineage societies the opportunity of agglomeration, for example in the case of war, this aggregation would dissolve once its goal had been attained. This principle based on kinship must be contrasted with the more durable principle of integration based on territory, or co-residence⁴. The Wessex communities seem to have succeeded in organizing polities 1000 km² in extent on this territorial principle⁵. Evidence of similar polities is very rare in Europe during the same period. The large territories which characterized the Wessex chiefdoms⁶ were not any more durable than chiefdoms documented in other contexts. This level of integration remained markedly intermittent during the Bronze Age. The available data, then, do not conform to the classical model of centralized territories. The data do support the centralization of power but only at a restricted scale and in three forms⁷:

l. A cluster of dispersed farms gravitate around a monument, a sort of tomb-sanctuary, which symbolizes the unity of the territorial community. This community is ruled by a chief who occupies one of the farms.

2. A cluster of farmsteads polarized by a village, near which is found the territorial sanctuary. These clusters should be composed of more numerous agricultural units than the preceding case (paradoxically the disparity of wealth is less evident).

3. Identical in organization to #2, but the central role of the village is held by a fortification. It appears that this type of settlement owes its existence to the control it exerted over long-distance exchange, especially over exchange in metal. This was a phenomenon of marginal zones where exchange between cultural complexes could take place outside the constraints of community laws. Each politically autonomous territory measured from 7 to 15 km in diameter during the whole period, except during periods of temporary expansion. Variation in territorial extent could be conditioned by other factors, of course – geomorphological, pedological, and topographic. All things considered, social stratification did not succeed in stabilizing Bronze Age polities. The polities existed but remained unstable. Thus, the mode of organization of Bronze Age societies retained many of the characteristics of the late Neolithic period.

Stabilization of political territories

During the Hallstatt B2/3-C periods (900-600 BC), the settlement pattern changes markedly. There is a great increase in the number of fortified sites. Small cemeteries of tumuli appear, often close to the fortifications. Typologies of ceramic and metal objects indicate the fragmentation of previous cultural units. Bronze hoards become more numerous – they are larger and their composition is more varied. Iron working becomes widespread. Rare earlier, iron objects increase rapidly in number during the ninth and eighth centuries BC. The phenomena noted above seem to imply social reorganization on a centripetal principle. A small fortification, the seat of the local aristocracy, polarizes each politically autonomous territory. The neighboring group of tumuli corresponds to a cemetery for the local aristocratic dynasty. During the same period, products from central Italy are beginning to be found north of the Alps. Most of the products of this region, a region undergoing urbanization, had traveled through the eastern group of cultures of the North-Alpine Complex. Baltic amber took the same routes in the other direction. The opulence of the eponymous cemetery of Hallstatt during the eighth and seventh centuries BC is explained not only by the exploitation of the salt deposits at the site, but also by its location near the most important east Alpine passes. It is in Bavaria, Franconia, and Bohemia that a dense concentration of tombs which produced horse trappings has been found⁸. Chariot burials are especially numerous there during this period⁹. Hallstatt is not unique in its intensive exploitation of salt. The majority of rock salt beds began production during this period. The marshes of the Seille in Lorraine are a good example. Excavation has revealed that certain artificial islands composed of the residues of salt domes belonging to this period were up to 15 m deep. The volume has been estimated at 3 million cubic meters. This implies specialized mining communities. The same characterization also applies to certain copper mines in upper Austria. There are also indications of economic specialization in animal herding. In Switzerland, lakeside stations practiced animal herding dominated by cattle rearing, while others specialized in sheep rearing. At Choisy-au-Bac in the Paris basin, the proportion of pigs in the faunal assemblage reached 60 percent with females predominating – an indication of particularly intensive specialized breeding¹⁰. Economic specialization increased not only at the inter-community level, but also within each community. The forge took its place beside the bronze workshop. Iron working requires the availability, knowledge, and expertise of full-time specialists. The presence of another kind of specialized craftsman – the carpenter-wheelwright – is evident in the ceremonial chariots. These craftsmen must have collaborated closely with the ironsmiths. It is during the same period that sophisticated looms first appeared. These allowed the execution of twilled weaving, a more sophisticated weaving technique which employs chevron and lozenge motifs of varied colors and of great delicacy; in short, they made possible the production of luxury cloths and tapestries. The existence of other craft specialists is implied by these great vertical looms. It should be emphasized that these craftsmen depended on the aristocracy, as the principal, if not only, consumers of their products. We must wait until the 6th c. BC to see iron play a significant role in tool kits; up until then, it was used primarily for the production of weapons. During Hallstatt D times (600-450 BC), contact with the Graeco-Etruscan world was instrumental in bringing about important changes in the social organization of west central Europe. During the two preceding centuries a north-south axis of exchange gradually superseded all other axes. This involved, principally, the eastern part of the North-Alpine Complex. During the 6th c. BC, the demands of the Mediterranean cities were increasingly felt in central Europe – but henceforth, in the western part, in association with the foundation of the Greek settlement of Massalia. The chiefs that were located directly on the principal communication routes obtained more Mediterranean diplomatic gifts. In the framework of a prestige economy, they gained in power, reinforcing their control of exchange with the Mediterranean civilizations, monopolizing redistribution, and finally subjugating neighboring chiefs. In this manner, centralized political units of a scale previously unknown in Europe were formed. These princedoms disintegrated in the 5th c. BC, whereas the intermediary role seems to have been transferred to the Tessin community on the one hand, and to the communities situated at the north-western periphery of the area held by the princedoms on the other – the Hunsrück-Eifel, Aisne-Marne, and Berry regions. This competition destabilized the “princes”, whose power remained very fragile because it depended totally on their contacts with the exterior. The zones in which diplomatic gifts were henceforth concentrated offer a contrasting image. They manifest structural analogies with the princedoms; with the exception of the region of Bourges (fig. 2), these cultural groups do not seem to be as centralized. Perhaps a political crystallization of the princedom type would have been produced if they had had more time to develop, as will be discussed later in this chapter. In the Hunsrück-Eifel, Aisne-Marne, and Berry regions, the density of habitation sites and cemeteries reached an unprecedented level during this period¹¹. The same conclusion is reached by a study of the settlement pattern in Bohemia¹².

Celtic expansion

From the beginning of the 4th c., numerous Celtic groups, organized under the authority of aristocratic chiefs, were established in northern Italy. The first, the Senones, probably came from Champagne. Another large contingent, the Boii, arrived later from Bohemia. This movement affected the whole of the Celtic world. Direct sustained contact with Latin and Etruscan civilization led Celtic groups to adapt by adopting not only a new mode of artistic expression, the famous Celtic art style of the La Tène period, but also, apparently quite rapidly, an urban-type territorial organization¹³. This evolution toward increasing organizational complexity is observed not only in Italy but also in the south of France. Small fortified cities became common in the fourth and third centuries BC. It does not appear, however, that the Celtic expansion in southern France was massive. Infiltrations here must have produced progressive Celticization without affecting the continued process of increasing social complexity, supported by the presence of Marseilles. Celtic expansion also took place toward the west and the east.

However, these disturbances led to the disintegration of the traditional exchange networks. In effect, prestigious Mediterranean objects only rarely reached central Europe¹⁴. Rich tombs became rare, and population density fell noticeably. This regression should not be exaggerated, however. Only the highest level of social integration disappeared. Local-level communities probably rediscovered their political autonomy but their internal organization remained stratified. Differences in social rank persisted, as is indicated by the variation in wealth found in tombs. The social range represented by this variation was simply more limited.

The density of cemeteries remained high, one every 4 km on average, but these were small, containing the remains of two or three aristocratic families at the most. The Duchcov votive deposit (Czechoslovakia) shows the development of craft production on a large scale for certain goods, such as ornaments¹⁵. Intra-Celtic stylistic homogeneity remained very strong, demonstrating the continuing intensity of exchanges throughout the cultural complex. With the occupation of northern Italy, the south-north exchanges also had a tendency to take an intra-Celtic form. It appears that Celtic expansion attained its maximum extent in the 3rd c. BC. While it is certain that these movements of Celtic people took place, it is unlikely that they were as massive elsewhere as they were in northern Italy. Regions such as Armorica and England, for example, retained a strong individuality and it is difficult to discern the influence of immigration and acculturation there.

Territorial reconstruction

Celtic expansion ended in the second half of the 3rd c. BC during the La Tène C1 phase. This is when some additional new elements appeared alongside older characteristics. The three most visible new elements are the adoption of coinage, the reappearance of the rite of cremation, and the creation of new sanctuaries. All three were still only in an incipient phase. Coinage, which remained rare and retained its intrinsic metallic value, did not circulate in a manner any different from the other prestige goods. The gradual nature of the transformation of the funerary rite suggests that the new elements should be explained in some manner other than by recourse to rapid and large-scale population movements. The new communal sanctuaries could signify the need of these sodalities to define their territorial boundaries more clearly and, by association, to strengthen community identity.

This transformation is associated with the La Tène C2 period, which began around 180/170 BC. Money came into general use as currency for the first time – its value a convention guaranteed by the issuing authority. Cremation was practiced almost without exception. The settlement pattern changed. Large boroughs that were centers of craftwork and marketing activity were established. Some of these boroughs erected enclosing ramparts – e.g. those at the site of Amboise¹⁶ in France. This recently excavated town seems to invalidate the supposed earlier development of the oppidum in the eastern Celtic area, with the sites of Stradonice and Pohanska cited as especially early examples. These hillforts, however, remained very few. In the present State of knowledge, unenclosed settlement agglomerations of tens of hectares were more numerous in the east as well as in the west – e.g. Mistrin, Střelice, Vienne, Bad Nauheim, Breisach-Hochstetten, Sissach, Basel-Gasfabrik, Feurs, Levroux-les-Arènes¹⁷. Several of these boroughs have produced evidence of coin production. All have yielded evidence of very specialized craft production in gold, bronze, iron, glass, bone, or pottery. Significant quantities of Roman amphorae are often found in them. On the whole, these sites possess most of the features that have been found in more recent oppida sites, with the exception of fortifications.

In the course of La Tène D1, some of these larger settlements were surrounded by ramparts. Occasionally these fortifications enclosed an area which exceeded the boundaries of the residential part of the site – i.e. Manching, Berne-Engehalbinsel, and Besançon. In most cases, however, the populations were moved to a nearby fortified prominence. Three examples are now well known – at Breisach, where settlement shifted from Hochstetten to the Münsterberg; at Basel, where the inhabitants moved from the Gasfabrik to the Münsterhügel, and at Levroux, where the Arènes was abandoned in favor of the Colline des Tours. Otherwise, as J. Collis¹⁸ has stressed, the majority of oppida are new towns, founded in extenso on sites devoid of pre-existing fortification. We must, then, assume a transfer of people and functions.

This process appears to have been gradual in spite of the fact that it seems to represent only a shift in topographical preference for settlement location (fig. 3). Large agglomerations were formed little by little by centralizing craft production and commercial functions. Those sites which produced their own money were also centers of a political power capable of ensuring its value. Some settlements had been fortified by the beginning of La Tène C2. Their ramparts already enclosed large areas on prominences – 80 ha at Stradonice and 50 ha at Amboise. Their internal organization is not well understood, which makes their interpretation difficult. A more numerous class of settlement was the unenclosed lowland borough. These evolved in two ways; either the original town was eventually fortified, or the settlement was moved to a nearby fortified prominence. This phenomenon was particularly prevalent during La Tène D1 and at the beginning of La Tène D2, that is, during the first third of the 1st c. BC¹⁹.

Were these fortifications a response to an insecure climate? It is possible that military conflicts took place between groups undergoing political restructuring, and opening up to Mediterranean trade. The negligible military value of certain ramparts should be noted, however, particularly the most extensive examples which were also the most costly. It is more likely that these were an ostentatious manifestation of power and a symbol of territorial control, a symbol reinforced by a dominating topographical location. The rampart was thus the principal public monument. Its length was a function of the power of the State, proportionate to the surface area of the polarized territory extra muros. The internal organization of these sites challenges the idea of a gradual process of centralization. The density of structures was actually very low. All oppida are characterized by household units composed of individual houses plus ancillary structures (granary, cellar, pit) centered around a palisaded courtyard. This household cluster evokes, in reduced form, contemporary farms. Thus, the traditional architectural organization was still the structural basis of the later settlements. To this were added open spaces, situated inside the fortification. These could serve as pasturage or for agriculture, which would decrease even more the contrast between rural and urban space. The oppida of temperate Europe can, however, qualify as towns. We are certain that several of the largest had inhabited areas of 20 to 40 ha, which even with a low settlement density implies a large permanent population. Various service activities were concentrated there and coinage was being produced. The oppidum was, thus, the seat of political and economic power. It tended to be situated in the center of the territory it controlled. This structural link between the urban and rural populace is manifested by the correlation between the surface area of the central site and that of its territory. This relationship is explicitly indicated by Caesar. It is testable archaeologically by the distribution of coin types produced by the center (fig. 4 and 5).

Reinforcement of social stratification

During the last two centuries BC, in spite of a limited amount of recoverable data, an uneven distribution of wealth is apparent among the known tombs within several culture areas – particularly in Belgian Gaul from Normandy to the middle Rhine. There, on the north-western periphery of the Celtic world, we know of thirty or so tombs containing parts of wagons, which were exposed, like the deceased, to the flames of the funeral pyre. Some Roman vessels have been recovered from these rich tombs as well. Concentrations of luxurious tombs around oppida are conspicuous (fig. 6). They resemble the configurations of “princely” tombs and settlements from the end of the early Iron Age. The evident hierarchy of fortified sites also expresses marked social stratification. The archaeological evidence does not contradict the literary sources. These distinguish three social categories: aristocratic warriors, from whose ranks were recruited the sovereigns or supreme magistrates; the druids, some of whom we know were aristocrats; and all the others – the majority of the population. The written sources also suggest that, at least for royalty, social status was inherited through the male line. For the highest social categories, filiation was patrilineal and residence patrilocal, as is suggested by examples of inter-tribal marriages where the spouse goes to live with the husband²⁰. At the other extreme of the social scale we can assume the existence of a slave category, in spite of the silence of the ancient sources on this subject²¹. We are uninformed as to their numeric importance and their function. In addition, we do not know if the servile class consisted only of domestic slaves, or if it constituted the base of the workforce as it did with the Romans. Social differentiation was not only vertical. Economic specialization was strongly accentuated, particularly in the towns. Many more individuals practiced craft production and trade full time. But of course, the vast majority remained peasants who produced the necessary surplus to supply the town-dwellers.

Internal factors

The archaeological record of the Bronze Age, in particular the social correlates of settlement patterns, suggests that for over a millennium the social hierarchy remained founded on long-distance trade and not on the control of land. Chiefs probably exercised control over the apportionment of land through their role as arbitrators of conflicts such as intra-community border disputes. They did not, however, control the primary products of the land and its surplus, quite simply because these only travelled short distances. For most local chiefs, secondary products with a higher exchange value were exotic – their control of these could only be weak and partial. Thus, they remained deprived of the economic base necessary for the permanent expansion of their territorial power²⁵. It is apparent that Bronze Age communities displayed expansionist tendencies²⁶ involving a significant demographic increase, the cultivation of previously uncultivated lands, and an increasing density of occupation. For reasons difficult to understand, the North-Alpine Complex spread more than others at the expense of neighboring complexes. Migration of settlers is one possibility. This expansion seems to have come to a halt at the dawn of the last millennium BC. There are several lines of evidence that suggest this. These can be interpreted as the result of a social and economic crisis – of an internal contradiction between a fast-growing population and stagnating modes of production – a situation which generated increasing conflicts.

The profound changes which appeared with the Hallstatt B3/C periods can be interpreted as the solution adopted to resolve the crisis, whereby autonomous communities would be stabilized and consolidated. The new strategy can be deduced from the increase in fortified centers, the development of an iron metallurgy that was less dependent on trade than tin-bronze metallurgy, the economic specialization of certain sites in the exploitation of salt or pig farming, and the adoption of innovative techniques in textile production such as the weaving loom with four warp bars, permitting the fabrication of luxury clothing and tapestries. In other words, communities were endowed with new secondary products produced from raw materials which were more widely distributed. Social stratification could be crystallized, given that the local economic base, which was controllable by the chiefs, was henceforth potentially present everywhere.

Once the necessary local base was in place, contact with the Mediterranean stimulated a considerable enlargement of both the scale and the level of integration, leading to the emergence of the principalities of the West Hallstatt Zone. However, as during the Bronze Age, this dependence on exterior influences left entities of a certain size in a very fragile position. It was not, however, an interruption in the provision of prestige goods which caused the disintegration of these principalities, since Etruscan goods continued to pass beyond the Alps during the whole of the 5th c. BC. There is another possible external cause for the collapse of the late Hallstatt polities which will be proposed in the following section. The ostentatious manifestation of power was eventually transmitted to the north-western periphery of the former principalities of the West Hallstatt Zone. It does not appear that a comparable degree of stratification had the time to develop there. Internal tensions, described by the ancient texts and compatible with the archaeological record, were caused by overpopulation and social conflict. These tensions were not resolved by an increase in social complexity, but rather by the emigration of excess population. This horizontal movement of people resulted in a decentralization of the stratification process during the fourth and third centuries BC. On a scale of social integration, the early La Tène polities barely surpassed the local level – a situation equivalent to that which existed during Hallstatt B3/C, between 900 and 600 BC.

During the 2nd c. BC, local factors made possible the emergence of State formations. Archaeology has produced evidence of significant change in agricultural practices. This phenomenon has not yet been the object of the research it deserves, but it is certain that many tools and agricultural techniques appeared during this period. Agriculture during the 2nd c. BC involved methods which, for the first time, permitted production on a scale sufficient to support a relatively large non-agricultural population concentrated in large settlement agglomerations. It seems that in temperate Europe more complex methods of agriculture were necessary in order to produce this surplus than those that have come to light in either the irrigated regions of the Middle East, or in Mediterranean areas of dry polyculture. After the 2nd c. BC, temperate European communities were able to maintain the conditions that were necessary to support a level of social differentiation as developed as that of the State. These internal developments, combined with external influence, enabled a breakthrough to a higher level of social complexity.

The dynamic of the Mediterranean world-economy

During the Bronze Age, displays of wealth and power (monumental tombs, rich grave goods) are frequently associated with exotic prestige goods of metal and amber. From this we can deduce that wealth and power were tied to preferential access to prestige goods. But this observation does not inform us as to whether this access was the cause, or only the consequence of social stratification. Another fact – the geographical localization of the most spectacular displays of power – tips the scale in favor of the first proposition. These displays did not occur close to the primary source of the critical material, but instead along the communication routes through which the prestige goods passed.

The Mediterranean world-economy integrated the North-Alpine complex during Hallstatt B2-3/C. The Greek and Etruscan towns experienced an increasing demand for raw materials which led them to enlarge their supply areas until they embraced a large part of the continent. In this vast exchange system, certain well-positioned local chiefs played the role of privileged intermediaries. They were able to monopolize trade and exchange, and controlled the supply of Mediterranean prestige goods, ultimately extending their influence into neighboring territories. They reduced local rulers to vassal status. These “princes” played the part of necessary intermediaries between the Mediterranean cities and the supply communities which they controlled on the one hand, and the more northern communities on the other: The economic system in which the Celtic “princes” played the role of intermediaries corresponds to that which F. Braudel²⁷ brought to light for the fifteenth and sixteenth centuries AD, which he called a “Mediterranean world economy”. I. Wallerstein²⁸ developed a model of the same type that covers a much longer time period.

I have recently developed an explanation of the sociopolitical disintegration of Hallstatt D in terms of a functional breakdown. The late Hallstatt principalities would have been stripped of their preeminent role by competition from emerging exchange routes²⁹, the middle Rhine, and the Champagne areas. In intruding into the preexisting exchange system, the negotiators from the Tessin could have looked for intermediaries further to the north. But the chiefs of these regions were no more than links in the chain of a down-the-line trade network, assuring contacts with their social equals and with the more northerly zones in order to assemble indigenous products in the form of metal, salt, meat, skins, furs and slaves. A more elegant explanation can be proposed: it can be argued that the scale of social integration was extended into new zones where ritual feasting apparatus was also present, but that it did not have time to crystallize because the invasion of Italy disrupted the exchange routes. According to this hypothesis, the Golasecca culture which occupied the Tessin³⁰ is seen as a functional outgrowth of Etruria in the second half of the 5th c. BC. It would have corresponded to an expansion of the first sphere of the concentric system of the world economy. This hypothesis requires the existence of towns in the Po plain at the fringe of the Golasecca culture. A representative site is the Etruscan center of Forcello at Bagnolo S. Vito, near Mantua³¹. Consequently, the transfer of the role of intermediary in the exchange network to Berry, Champagne, and the middle Rhine could represent a variation of the second sphere of the world-economy.

The adoption of an expansionist strategy by the northern Celts interrupted this process. It is not known why this solution to internal tensions was favored over reinforcing the existing level of integration. Clearly, this choice had important consequences for the structure of the world-economy. Nevertheless, the latter was not totally destroyed. The evidence suggests that the object of this Celtic expansion was to encircle the power centers of the world economy – to control the whole of the second sphere, after having attempted to advance upon the cities of the first sphere in order to profit more directly from the system. This probably would not have been a problem if the Celtic expansion had taken place primarily in the intermediary zone. However, by attempting to gain too much too quickly, the groups which adopted this strategy ended by damaging the integrity of the Celtic world as a whole. In the course of La Tène C2, from 180/170 to 130/120 BC, the whole of the first sphere of the world-economy fell under Roman control. The future imperial capital at first was the center of a system which was being revived. The second sphere corresponds to the Celtic States zone where oppida began to appear gradually. All excavated oppida have produced evidence of intensive commercial interaction with Rome. Celtic coinage, inspired by Mediterranean prototypes, is one of the earliest indications of the restructuring of the world-economy. The concentric functional organization consisted of three levels of politico-economic complexity, decreasing from the center to the periphery.

Because of the intensification of exchange with Rome, however, the difference between the levels of development of the first two spheres was reduced. The beginnings of the princely phenomenon appeared in the third sphere. When Roman power extended as far as the Rhine, integrating the whole of the Celtic States zone, true principalities formed in free Germany. This phenomenon is evidenced, above all, in the famous group of Lübsow tombs³². Roman influences spread from 200 to 600 km beyond the Limes in close association with Roman trade. As in the 5th c. BC, the growth of the first sphere caused the displacement of the second sphere further from the center.

Theoretical implications

It is of interest to conclude by submitting our case study to further theoretical analysis. First, it appears that the evolution of the Celtic world does not conform to frequently proposed explanations which assume that there was a growing relative scarcity of resources³³. According to this theory, scarcity causes conflicts, the resolution of which leads to a delegation of power by a populace to an arbitrator. Besides, a scarcity of resources does not necessarily give rise to an increase in social stratification. The emergence of the State in the Celtic world was the result of an intensification both of agriculture and of long-distance exchange. This explanation corresponds most closely to the neo-Marxist explanatory framework proposed by J. Friedman and M. Rowlands³⁴, which is as appropriate to the evolution of the State in Mesopotamia as it is to similar developments in Peru or China. The intensification of agriculture allows for the support of a rise in non-agricultural production and this entails a cumulative process of centralization/stratification. Clearly, the more production intensifies, the more frequent are the occasions for dispute and, consequently, the greater the necessity for arbitration³⁵.

A solid local economic base is indispensable to the emergence of a more complex, enduring organization. In the case of principalities such as those of the Celtic States, the influences emanating from more complex neighboring societies stimulated the process of social stratification. However, this influence only reinforced and accelerated a phenomenon made possible by internal factors and by technical progress which allowed the intensification of production.

It can be argued that innovations in technology came about in response to the well-known contradiction between population growth and the carrying-capacity of the environment³⁶. The economic foundations put in place in the ninth and eighth centuries BC were, however, incapable of supporting a political scale of integration greater than tens of square kilometers. They did permit the stabilization of territorial units and social hierarchies, which had been fundamentally unstable up to that point. During the 6th c. BC the considerable increase in the level of integration, due to the functioning of the world-economy, was just as artificial as the unsuccessful attempts of the Bronze Age. The principalities seem to have disintegrated because of a modification in the spatial distribution of economic functions at the heart of the world-economy (concentric, functional tripartition). This occured as a result of external causes. Significantly, the disruption of the south-north trade networks in the fourth-third centuries BC brought about a return to the scale of integration which had existed from the ninth century BC onwards. In order for a superior level of social complexity to develop, a change in agricultural production was necessary. Techniques capable of intensifying production on the heavy and deep soils of temperate Europe were required in order to deliver a surplus sufficient to supply the needs of a greater number of non-agriculturalists.

The evolution of Celtic polities presents another interesting phenomenon relevant to a more general discussion of social stratification. An increase in social complexity is not the only means of remedying internal tensions. Territorial expansion or emigration are others, as is suggested by the aborted processes at the end of the 5th c. BC. At that time the level of integration decreased, showing that the process is not irreversible. Thus, we can determine the necessary conditions for the emergence of the State, but we are unable to determine their extent. This difficulty may be due to the fact that there is a certain degree of probabilistic chance involved in the choice of a new type of social organization. The process of evolution acts in the manner of Prigognine dissipative structures, as suggested by van der Leeuw³⁷. Differential access to resources is generally considered a necessary condition for social stratification. In all known cases, this inequality has preceded the formation of the State, but it has always increased with its consequent development. G. Johnson³⁸ has proposed that social hierarchies are the result of a differential capacity for processing information. These two points of view are not incompatible. Privileged access to material resources presupposes the processing of information concerning the location of resources, the conditions of their transport and distribution and, above all, the partners and the codes of exchange. It is not so much that the elites are more capable of procuring the material goods, but that they have the means of disposing of these goods. What the elites exchange are agreements which ensure the supply of goods at a precise location and time³⁹. The processing of information is first and foremost a service activity. The crucial importance of information processing is sharply delineated in particular in the institutions which characterize the State. In effect, the State is defined as a form of government endowed with a specialized institution for the processing of information: an administration. Thus, it should be possible to see our problématique in terms of self-organizing systems theory. Technical innovation, the necessary condition in the subsistence domain, and its corollary, intensification of production, depend upon a series of steps involving the processing of information – innovation, diffusion, exchange of services, etc. This series of prestations makes it possible to overcome in a positive way the imbalance between population and the capacity for production. Population growth, which multiplies the number of potential parties, consequently increases the quantity of information to be processed. It can thus provoke scalar stress which, in the absence of sequential hierarchies, can be resolved either by hierarchization or by the creation of additional hierarchical levels⁴⁰. The contradiction can be resolved in a negative fashion by a reduction of population, which can be controlled to a greater or lesser degree by contraception, a rise in marriage age, emigration, war, disease, or famine. These observations and their implications suggest that we should integrate into our approaches methods adopted from the study of chaotic systems, systems in which evolution defies expectations. In spite of their complexity, these systems are not random. They possess structure. Physicists, meteorologists, and biologists have noted the preferential or “attractor” States which a system undergoes in the course of its evolution. In addition, these attractor States frequently possess a similar structure on various scales of observation – a fractal structure⁴¹. Clearly, this type of approach could be applied to human social systems – the ultimate complex systems. An examination of the evolution of social stratification in the Celtic world reveals a systole/diastole type of dynamic, a cycle of evolution and devolution. This phenomenon requires us to recognize the significance of agricultural intensification in the process of increasing social complexity and to equivocate the role of long-distance exchange, even in a system of the world-economy type. This statement, for all that, does not leave us in an inextricable knot of internal contradictions.