Oppida and social “complexification” in France

Local conditions

One must go back as far as about 400 BC to grasp the conditions behind the “complexification” of non­Mediterranean Gaul (fig. 1). The4th and 3rd c. BC were marked by considerable instability. Early texts and archaeological data indicate movements of population, from the heart of the north Alpine cultural complex into northern Italy, but also towards the east and west². Migration was caused by population creating too strong a pressure on resources, provoking as a result serious internal conflict. Study of settlement and cemetery distribution suggests in fact important population growth during the 5th c. BC and a decline in the 4th c. BC³.

Emigration in organised groups is one of two solutions traditionally employed to reduce tensions when these endanger a community’s very survival. One procedure is segmentation or fission. The other solution is to reinforce political power by setting up a hierarchy. It is difficult to understand why one solution was chosen rather than the other. Titus Livius⁴ tells how an Etruscan brought ‘Celts’ to his city to deal by force with a family conflict Pliny the Eider⁵ underlines the attraction of Italy and especially its wine for the ‘Celts’. These accounts should not be taken too literally, as they resemble myths. We can note however that emigration generally appears to represent an opportunity for individual and collective valorisation, with good perspectives for adventure, glory and enrichment. It thus seems likely that fission was the easiest solution, partly because the emigrants were keen to leave, and partly because the inhabitants of coveted lands were incapable of offering resistance, due to their own dissensions.

Despite the departure of part of the population, the distribution of cemeteries and the richest graves shows the persistence of two levels of socio-political integration. The cemeteries are 3 to 7 km apart, whilst the richest graves are 10 to 15 km apart⁶. We can thus envisage territories averaging 122 km², each including about six local communities. Sociological analysis of cemeteries shows, in addition, the existence of well differentiated social strata, the membership of which was transmitted along hereditary lines⁷.

Supra-local power was apparently based on control over external resources: copper, tin, amber, coral, weapons, ornaments, tools and vessels in bronze and iron, goldwork, and chariots. Elites nevertheless seem to have taken additional control over certain internal resources. These were rare but locally available raw materials (for example metals in the Hunsrück-Eifel), but also prestige goods which could be produced almost anywhere if the political will existed: salted meat with intensive stock-raising and a regular supply of salt, or luxury textiles with specific stock-raising and the development of very specialised skills. Worth noting here is the appearance, in the Aisne­Marne region, of batteries of several dozen large storage pits on certain sites. These are perhaps signs of increased control over food supply by political leaders. The general picture is thus of a specialised and hierarchical political organisation, with a limited scale of integration, since a return journey from the centre to any part of the territory could easily be made in one day. The economic subsystem seems to have been quite substantially controlled by the leaders.

By choosing emigration to reduce population surplus, the communities to the north of the Alps were able to continue and even accelerate their high population growth. Then, as always happens, when expansion stopped the birth-rate would not immediately have adjusted to the new situation. Communities thus faced the same problem as before expansion, pressure on resources and internal conflict. Since fission was no longer a possible solution, breakdown of social relations could only be avoided by the establishment of a hierarchy.

As we have just seen, population growth was a crucial factor in the social phenomena that took place during the first half of La Tène. Population pressure clearly leads to pressure on critical resources. We can add that it also produces a rise in social density⁸, in other words more frequent communication between individuals and different groups forming the whole society. This results in a weakening of the capacity for processing information⁹. These two factors are responsible for an increase in social conflict. Threatened with disintegration, society has to find a means of easing pressure on critical resources and reorganising internal communication.

External conditions

I have set out my ideas elsewhere on the existence of a world-system centred on the Greek and Etruscan cities around 500 BC¹⁰. More recently, I have tried to show how this dynamic process lasted, despite the population movements of the 4th and 3rd c. BC¹¹. These movements can in fact be considered as extreme forms of attraction of the more active centres, even if they disrupted exchange for a while.

During the 2nd c. BC, Rome took control over the rim of the west Mediterranean basin. The whole of the first circle of the world economy, as it has been shown to have existed in late Hallstatt times¹², thus fell under Roman influence. The future imperial capital initially became the centre of the Mediterranean world economy, restored after “Celtic” turbulence. The second circle corresponds to the zone of “Celtic” states within which oppida gradually appeared. All these sites have produced evidence for trade with Rome. Celtic coinage, inspired by Mediterranean prototypes, was the earliest sign of the restructuring of the world economy. This functioned in concentric manner with a three-level hierarchy, the political complexity of which decreased from centre to periphery. The difference in levels of development between the centre and the second circle, however, was reduced in favour of the intensification of exchange with Rome. Premises of a princely phenomena started to emerge in the third circle. And when Roman power was extended as far as the Rhine to include the whole “Celtic” states zone, true princedoms formed in Germania libera. This phenomenon, expressed through rich burials, flourished between 200 and 600 km from the Limes in direct relationship with Roman trade¹³. As in the 5th c. BC, enlargement of the central zone caused the displacement of the second circle further away from the centre.

The world-economy model also fits the differences observed within the “Celtic” complex. Spread out on the periphery of the Mediterranean civilisations conquered by Rome, the “Celtic” complex, just before Caesar’s conquest, was itself organised in three concentric rings. In the south lay the most extensive centralised “Celtic” states. The territories of the Arverni, Bituriges and Aedui measure from 80 to 100 km in radius. This ring of major States was quite broad. Behind it were the great territories of the Lingones, Carnutes and Senones. It was this part of Gaul that was covered, from the Atlantic to Lake Geneva, by states which produced currencies exchangeable both with one another and with Roman dinars. Caesar used these states as a support for launching bis campaigns on more distant Gaul, a region he intended to exploit¹⁴.

Further from the Mediterranean, from Armorica to Champagne, lay smaller states, 30 to 50 km in radius. Although smaller, each had a capital where coinage was produced and they imported a fair amount of Roman products. Here, not far from the northern limit of oppida, from the Seine estuary to the Rhine, elite burials contained Roman vessels which served as status symbols, sometimes associated with parts of chariots taken from the funeral pyres¹⁵. Further beyond, from Britannia to northernmost Belgic Gaul, lie less developed States which are only reached by very limited quantities of Roman products. This concentric zonation, already noted by other authors¹⁶ was deformed by the main cross-routes where the traffic’s intensity provoked a greater social reaction; in particular the Rhone-Saône axis and its three branches: Seine, Meuse-Aisne-Somme and Moselle-Rhine. Thus, within each of the concentric zones, social ‘complexification’ falls off in stepwise fashion with distance from the centre.

The model therefore appears very similar, even in scale, to that of the world economy of the 6th and 5th c. BC. The main difference involves the level of social complexity reached by the second circle. The “Celtic” states which had relations with the Roman world had already crossed the threshold of urbanisation and state formation. It is true that the Roman republic represents a type of state which was economically and administratively more complex than earlier Etruscan and Greek city-states. However, the gap between centre and semi-periphery had been reduced.

Long-distance relations with the Mediterranean States clearly stimulated the Gauls. It remains difficult however to measure the degree of stimulation. Influence could, in fact, have been very direct through the intermediary of Cisalpine Gaul. The available evidence from northern Italy is inconclusive, but urbanisation and state formation could possibly have occurred earlier near the Etruscan and Latin States¹⁷. Adopted with various adaptations in the Cisalpine laboratory, these changes could have been rapidly transferred into Transalpine Gaul. Yet at the same time, a social mutation of this importance could not have had such a fast effect on societies insufficiently engaged on the path of “complexification”. Consequently, internal and external factors logically combined in this zone which had been located for five centuries on the periphery of state societies.

The economic subsystem

To respond to pressure on resources, more must be produced. This means increasing the surface area cultivated and/or innovating with new farming practices and tools¹⁸.

Palaeobotanical data is still too scarce to date precisely the adoption of new crops or new production cycles in non­Mediterranean France. We know that at the end of the Late Iron Age several species of wheat, the staple plant food, were cultivated and we can suppose that there were two annual harvests¹⁹. We do not know however whether these changes took place in the 3rd or 2nd c. BC or earlier. Neither do we know whether or not the practice of two annual harvests was widespread.

Faunal data is rather more abundant, but we can do no more than note the considerable variation between sites in the proportions of stock raised throughout the Iron Age. This probably indicates a certain degree of specialization, but we cannot yet distinguish between chronological, geographical, or functional factors²⁰.

The settlement data is more explicit. The vast majority of structures known as fermes indigènes and the ditch systems with which they are often associated belong to the 2nd and 1^st c. BC²¹. These are field boundaries which reflect increased control over the landscape. The aim was to deal with herd movement more effectively, to protect crops from animals, to improve crop separation, and perhaps even to create clearer property boundaries. In short, production was rationalised to increase yield. This result was obtained in return for a substantial investment of labour, since the ditches and palisades had to be constantly maintained. During the same period, new and more effective tools appeared thanks to widespread use of iron: ploughshares, spades, hoes, picks, sickles, scythes, shears. The same concern to improve production is evident in the rapid adoption of the rotary quern²².

In sum, it was the way in which people modified their surroundings that best reflects the scope of changes affecting agricultural practices. The scars of this genuine agricultural transformation are still visible in the present day landscape. We can note that these multiple ditch systems not only imply rationalisation to intensify production, but also reveal a concern for greater material display of rights, or even property of land, which is a logical concern in a situation of population pressure.

The political subsystem

Social density, which should not be confused with population density, depends on a combination of several additional factors:

• the spatial distribution of population, settlement, and exchange nodes,
• the size and shape of the territory,
• the means of transport and information transmission.

These characteristics²³ combine to affect the degree of intensity of communications within society²⁴. To avoid the destructive effects of densification, a society bas to organise communication by establishing a hierarchical means of processing information and by improving transport and information transmission.

In intensively explored regions, settlement patterns indicate a rise in the number of sites, with increasing functional and hierarchical differentiation between them²⁵. The end of the 2nd c. BC is in fact marked by the emergence of large 5 to 10 hectares villages with varied craft and commercial activities, which can be termed small towns (bourgs)²⁶. Thus, rural population density was not only increasing, but rose to such an extent as to form agglomerated sites containing higher numbers of people than ever before in non-Mediterranean Europe. This phenomena required agricultural production capable of feeding the non-agricultural section of the population. We have seen that signs of considerably intensified production began to appear a little earlier. The supply of food to the inhabitants of the small towns also suggests a sufficiently developed transport infrastructure. Latin texts underline the quality of Gallic vehicles, which implies the existence of suitable roads.

These small towns were occupied for varying lengths of lime, and then were either fortified, as was often the case with riverside sites, or more commonly transferred onto nearby higher land and fortified. These two processes gave birth to oppida. It seems unnecessary to see the transfer to higher ground as the consequence of an archaeologically invisible catastrophe. We can consider the phenomenon more simply as a means of reinforcing social cohesion and political control during a period of change and, therefore, high instability. The ramparts enclosing oppida were often of poor defensive value. They were extremely long, running up and down slopes, and their facing did not resist war machines for long. Above all, the size and undeniable aesthetic quality of these ramparts, especially their monumental gates, displayed prestige and power. The substantial earthwork showed a capacity for mobilising and organising a large labour force, and, at the same time, its collective construction reinforced a sense of belonging to the social and political unit. The ramparts probably also served to delimit an area within which the political authority guaranteed the security and regularity of exchange.

In comparison with Mediterranean towns of the same period, oppida were not densely occupied. At Manching, outside the central zone where interpretation of the plan is hindered by superimposed traces of occupation, there are ten houses per hectare. If one nuclear family lived in each house, we can estimate the total population at around 5000, since 100 hectares of the 200 hectares enclosed by the rampart were actually inhabited. An estimation based on the meat consumed gave similar figures: 3400 to 5100 people²⁷. This relatively low density was nevertheless much higher and more permanent than ail the concentrations which may have formed in earlier limes. There resulted a substantial and rapid increase in the number of interactions between individuals and between groups, with the inevitable consequence of increasing the quantity and variety of information received by everyone. As Johnson²⁸ has clearly shown, this creates anxiety and organisational problems due to a limited capacity for dealing with information.

Society had thus become much more opaque, as nobody could immediately perceive the configuration of the whole²⁹. The most commonly adopted solution for this type of problem is to organise hierarchically the organs of information processing. The internal communication of a society, however, worsens when the number of hierarchical levels become too great, since the quality and quantity of information decrease with each transmission between levels. These two constraints are extremely constrictive limits for the political management of a society. To ease these constraints, innovations in communications and information processing have to be made. There is no doubt that, in ail emerging states, writing was the necessary innovation. It added the finishing touch to strengthening the means of government.

A period of high population growth and difficult fission, the end of the Late Iron Age was very logically marked by the reappearance of a third level of integration whose centres, the oppida-capitals, were on average 65 km apart and controlled secondary centres 10 to 30 km apart³⁰. Even greater political entities existed in central Gaul. These seem to have had four levels of integration with a scale of over 20 000 km². Caesar called these political entities civitates. According to the conqueror these were centralised and politically autonomous units. With few exceptions, the capital was the largest fortified enclosure in the territory. Some of these oppida were certainly producing coinage³¹, which was mainly used for trading within each autonomous political unit. This economic instrument implies a political organisation able to control all the coinage in circulation, exchange of currency at frontiers and the authenticity of coins.

The use of coinage is the principal argument for seeing the Gallic civitates that Caesar encountered during bis conquest as genuine states. The quite considerable strengthening of government can also be seen in the “Celtic” inscriptions transcribed into the Greek alphabet, increasing numbers of which have been found over the last few years along the Rhone valley, right into Burgundy. Bronze styluses and frames for wooden tablets probably covered originally with wax have been discovered as far away as on an oppidum in Bohemia³². One remembers Caesar’s evocation of tablets containing lists of names of Helvetic emigrants (BG, I, 29). According to Caesar, only Druids would have used writing to keep accounts and public and private registers. We can thus suppose the existence of a genuine administration run by the holders of religious legitimacy. In particular, they would have managed treaties and contracts³³.

A specialised government machinery had clearly emerged at this time. Caesar’s writing about the Gallic war tums out in fact to be quite explicit about the “Celts”, form of political power. This was institutionalised and incarnated in the person of a sovereign or a magistrate elected amongst the aristocracy, and exercised by networks of clients: the great aristocratic families. Internal conflicts due to competition for supreme power brightened the life of these governing families. Nevertheless, an embryo of administration was set up under Druid authority to ensure the stability of the institutions.

The ideological subsystem

Durkheim³⁴ termed social density, dynamic density and stressed a moral dimension represented by the degree of cohesion between social segments. We are all aware today that ideological cohesion is the best social cement. Numerous authors have stressed the integrative function of ceremonial activities and the fact that their intensification is often used to overcome a crisis in the social system.

Current evidence dates the first shrines clearly separated from cemeteries to the end of Hallstatt c.500-450 BC³⁵. From the first half of the Third Century BC onwards, shrines seem to have become more common. They consisted of sacred areas, delimited by ditch and palisade, where animal and human sacrifice was performed. These shrines contained trophies and tended to be located near the centre of territories and close to frontiers³⁶. We can thus discern their role as territorial markers, revealing a period of reinforcement of power. In relation to all members of the community and foreigners one had to express a will and capacity to defend one’s territory.

The threat of war indeed requires the community to control its land, which implies exclusion of strangers and foreigners. As well as physical protection, the exclusion of others allows the community to affirm its identity by underlining its differences, and this affirmation can go as far as a war of aggression. The twin function of territory – differentiation in relation to the exterior and reinforcement of collective identity within – is particularly important during periods of changing scales of integration. Incorporation of different communities into the same political unit in fact requires ideological integration of these differences, since force alone cannot maintain the amalgam. The whole system of values is modified and the new social norm bas to be interiorised by the members of the new entity for there to be consensus. Traditionally this is undertaken during communal religious ceremonies.

The founding of temples implies a priesthood, which was a further step towards specialization. Perhaps for the first time in the study zone, political and religious power were separated. Up until then the two powers had probably been held by the same person or group. The first service that elites offered to their community – or so it is believed – seems often to have been of magico-religious nature. It involved ritual control over fertility and communication with the ancestors and gods³⁷. The separation remains quite relative, however, as the priests carried out important administrative functions. For this reason, the phenomenon should not be seen as a reduction of power, but rather as a development of the personnel used by the ruling body for governing. In a society where the scale of integration and type of political organisation was changing. the priests diffused a message legitimising power and the necessary modifications in the system of values.

In sum, the archaeological evidence suggests the following chain reaction: first, parts of the ideological domain were modified in order to adapt the value system, then the economy underwent intensified production, and finally there was a political trend towards strengthening the means of government As soon as it was affected by change, each subsystem worked retroactively on its counterparts and this produced a cumulative chain of effects.

A slow and discontinuous process

Ethnographic studies have shown the omnipresence of the supernatural in tribal systems and its monopolisation – as a condition of imaginary production – by the elites in chiefdoms and archaic States⁴⁰. The same idea is expressed by Godelier⁴¹, for whom the first forms of state developed more through general cooperation than through minority violence. He deduces from this that the exercise of power should appear as a service rendered by the leaders, and underlines that it is usually of magico-religious nature. The leaders also believe in the usefulness and even the indispensability of the service. As a result, it becomes social reality, even though it is an illusion in absolute terms, and creates for the dominated:

“une dette qu’ils doivent honorer par le don de leurs richesses, de leur travail, de leurs services, voire de leur vie”[…]“La voie est alors ouverte pour que le pouvoir dû à la fonction devienne un pouvoir d’exploitation, et pour que le travail en plus mis au service de tous devienne du surtravail⁴²”.

“a debt which they must honour by the gift of their wealth, their work, their service, and even their life”…“The way is thus open for power from function to become power for exploitation, and for extra work for the benefit of ail to become overwork”.

In Europe the process seems to have been quite gradual. From the beginning of the Bronze Age, we find privileged individuals apparently holding politico-religious power and controlling exotic resources. There was a tendency for this form of concentrated, specialised power to become hierarchical. Political and religious functions thus became increasingly separate. Now at the beginning of the first millennium BC, bronze, the most important prestige good, was being used for agriculture and crafts. As a result, elites controlling bronze began inevitably to control the means of food production as well. This means that local communities governed by kinship relations started to lose control over the subsistence economy. The phenomenon becomes more pronounced at the beginning of the Iron Age, with not only the widely adopted new metallurgy but also the large-scale exploitation of salt, luxury textile manufacture, and specialised breeding of certain animals. The distribution of resources thus became less unequal, since more local communities were able to produce prestige goods. In early La Tène batteries of dozens of large silos occur on certain sites, suggesting a form of control over food surpluses by supralocal elites.

In other words, over about two thousand years, we see a type of society in which the political. ideological and religious domains were so embedded in the socio-kinship domain that they were more or less imperceptible evolving into a type of society where organisation became more diverse, its functions less interwoven. They could now be more clearly identified. They remained interdependent but became relatively autonomous of the socio-kinship subsystem. Its only function now was the maintenance and reproduction of society, principally by regulating kinship relations. This process takes place in societies which can be classed as chiefdoms. As such, even before the adoption of a state form of organisation, kinship relations had ceased to unite all functions of social life. This should not make us think in binary fashion e.g.:

Rather we should see a more gradual development, patterned by numerous moments of change, towards increased integration but sometimes towards disintegration too.

“Complexification” as resistance to entropy

It has been useful to conceptualise the proposed scenario along systemic lines. This has also enabled us to approach the questionof the apparent determinism of “complexification” and its underlying principles. The social system as I see it is composed of six subsystems (fig. 2). The psycho-physiological is the motor of population growth. This is limited by the ecological subsystem, which has a given carrying capacity. Demographic regulation generally takes place when the environment is threatened by over-exploitation.

Now if ecological conditions become more favourable, the traditional form of regulation loses its ritual character, and this leads quite quickly to pressure on resources and an increase in social density. In this situation, society can bring traditional methods of regulation into play; this solution must have been adopted in most cases. Society can also innovate in order to intensify production. Such an action on the economic subsystem causes a new rise in social density, and this ultimately requires an adaptation of the system of values – in other words the ideological subsystem – as well as an inevitable strengthening of the means of government – in other words a change in the political subsystem. The socio-kinship subsystem, which is dominant up to then, is now affected; the role of kinship relations is reduced. Depending on the situation, the order of succession of these four factors could vary. But this is quite unimportant, since they are interdependent. Change in one changes the others and all retroact with each other.

One condition of change recurs in the sequences recorded: a rise in social density. This variable produces both conflict and the correlative necessity of ideological, political and economic integration. Thus, it seems as though social ‘complexification’ was determined by social density and was consequently destined to become more pronounced. We note in fact that societies ‘chose’ socio-economic innovation in at least six independent cases: Middle East, Indus, North China, South China, Mesoamerica, and Andean America These original Neolithic centres have ail generated locally or in the neighbouring areas strong states and their form of organisation was diffused and has been adopted almost everywhere at the present time. The same is true of the state form of organisation. This signifies that history can be random, circumstantial and unpredictable in the short-term and on the micro-social level, but seems determined by a tendency towards “complexification” in the long-term and on the macro-social level.

This tendency could be explained by the fact that “complexification” is necessary for the maintenance of stability in societies with high social density⁴³. Cohesion and order are maintained here at the cost of energy derived from a difference in potential resulting from the establishment of social hierarchy and from intensive exploitation of the environment. Analogous to the working of thermodynamic machines, this principle is characteristic, according to Lévi-Strauss⁴⁴, of societies which he terms ‘hot’ for this reason. As social conflicts tend to result in general levelling-out, he adds, such societies look for other kinds of disparity, which they then re-establish through colonialism or imperialism. This is precisely the sort of mechanism described by Wallerstein⁴⁵ and Braudel⁴⁶ for modem limes, and which I am proposing to apply to the European Iron Age.

We can consider that heavy pressure on resources and arise in social density, reduced for a while by resorting to emigration, led societies to reinforce the function of the economic, ideological and political interdependent subsystems to bring about integration and thus maintain social bonds. The process was stimulated by contacts fostered with the Mediterranean States. The latter probably also provided ideological and political models. In the economic field, however, total innovation was necessary, since Mediterranean agriculture could not be adapted to more northern zones.

Finally, it can be suggested that social complexity is necessary to maintain stability when social density increases. The aim would be to create a social differential, like thermodynamic machines which work by shifts in temperature. Admittedly the second principle of thermodynamics States that entropy, disorder and equality inevitably prevail. But this does not invalidate our point of view on growing complexity, since the social subsystem makes the effort to complexify at the cost of increasingly rapid environmental exhaustion. In general terms, the process does tend towards entropy and so is not in contradiction with the second principle of thermodynamics.