From the Hallstatt to La Tène period in the perspective
of the Mediterranean World Economy

The data

The transition from Hallstatt to La Tène is dated to around 450 BC. This typological break corresponds to the earliest known reference to the Celts². The Celts’ territory was very plausibly equated with the distribution of La Tène material culture. Having explained the 5th c. BC change as the result of migrating Celts, attempts were made to show that they arrived in the north Alpine region first. The geographical superimposition of the La Tène culture on the Hallstatt culture was unquestionable. Migration theories persisted, however, and it was thought that prior to La Tène only the aristocracy in this region was Celtic. But here again the difficulty was identifying the geographical origin of the invaders.

More recent studies undertaken on the raw data have shown clearly that material dated to La Tène A resulted from local evolution. Thus, the Hunsrück-Eifel³ and Aisne-Marne⁴ groups start in Hallstatt D. The transition is equally subtle in artefact typology in eastern Switzerland, where burial rites gradually change in the course of La Tène A , from secondary mound burial to flat inhumation⁵.

These observations are totally incompatible with an immigration hypothesis.

Re-examination of the basic documentation also permits refinement of chronology and recognition of polythetic cultural entities. As a result, the changes occurring within the north Alpine context from Hallstatt D to La Tène A can be defined more clearly. These changes are structural rather than stylistic in nature. Hallstatt D princely residences like Vix⁶ or the Heuneburg⁷ are no longer occupied. The very rich tombs, containing Greek and Etruscan imports, which are grouped around the fortified residences are no longer built⁸. In La Tène A , display of wealth in burials occurs further north, mainly between the Rhine and Moselle⁹. Mediterranean imports follow this geographical shift¹⁰ (fig. 1). The phenomenon cannot be explained by a decline in the commercial activity of the Greek city of Marseille, since Mediterranean products, most of which are Etruscan, continue to cross the Alps. It is therefore necessary to seek an explanation in the internal functioning of Celtic society.

Sites identifiable as princely residences are fairly regularly distributed in a zone stretching from Burgundy to Württemberg. It has been suggested that each site constitutes the organising centre of a territory of about 50 km in radius¹¹. The strength of this hypothesis is reinforced by analysis of the spatial distribution of certain artefact types¹². This kind of analysis enables a hierarchical classification of the observed cultural entities to be undertaken, along the lines proposed by D. Clarke¹³. Thus, each princedom seems to represent a group, and the whole princedom zone a culture.

It is generally agreed that a Mediterranean stimulus underlies the development of princedoms. I have suggested that if social reaction to Mediterranean influence was rapid, this was because ‘barbarian’ society already possessed a centripetal territorial organisation from the 9th c. BC onwards¹⁴.

Several local chiefs could have monopolised the function of intermediary for providing the raw materials that were required. In return they received prestigious gifts which they redistributed according to a model that is now well-known¹⁵. These observations fit the Mediterranean world-economy model elaborated by F. Braudel (1979) for 15th and 16^th c. in Europe. In the Hallstatt period, a vast European system is discernible, hierarchically organised in three concentric rings. The Celtic princedoms control a dynamic north-south exchange axis in the intermediate zone.

A tentative interpretation

The Hallstatt D princedoms collapse and burials with Mediterranean imports appear on their northern borders. It is difficult to imagine that the two events are not causally linked. Our evidence is, however, not compatible with the idea of a mass geographical transfer of the princely phenomenon. The spatial distribution of rich Marne-Moselle Culture graves differs from Hallstatt D princely burials. The former is widely distributed without marked concentrations, and less luxurious, lacking the complete banquet services which characterise the latter. No settlement comparable to the princely residences have yet been discovered, despite the fact that this is one of the easiest types of site to recognise. The spatial model evokes a network of juxtaposed territorial communities. Certain individuals are apparently wealthier than others, but nothing suggests that they subordinated their less wealthy counterparts. In short, if there was a transfer of mediating function, it did not have the same social effects. One can thus assume that only a part of the intermediary function previously monopolised by the princes was involved in this region. For this hypothesis to be acceptable, it is necessary to investigate who carried out the remainder of the function in the Mediterranean world-economy.

The Rhône corridor certainly loses importance in the 5th c. BC¹⁶, very probably due to competition from the Black Sea colonies founded by Athens, now the hegemon of Greece. This leads us to examine the Etruria-Middle Rhine axis. The distribution map of chariot burials¹⁷ shows few examples in the Ticino region. Containing rich grave goods, they date from the 7th to the 5^th centuries BC. The community occupying this very favourable passage across the Alps probably benefited from north-south trade during the Hallstatt period. By the end of the 5^th c. BC the community, with its uniform material culture called Golasecca, shows increasing wealth through its burials. Imported vessels, in particular Schnabelkannen, are very frequent¹⁸. This community even manufactures its own bronze vessels, the Ticino situlae¹⁹. The relations between the Ticino and the Middle Rhine are illustrated by the presence of concentrations of Rhineland imitations of Ticino situlae. The relaying role played the Golasecca Culture is also shown by the distribution of its own products north of the Alps²⁰.

All the evidence suggests that this Alpine community interposed itself between Etruria and the Celtic princedoms. By establishing control over trans-Alpine exchange, it robbed the princes of part of their monopoly. As a result, the princes received fewer prestige goods and lost their redistributive capacity. Their power crumbled. North-south trade no longer passed through the relays which they controlled. It seems very significant that the new relays were positioned at the entrance and exit of the Celtic Complex: in other words, on the border between the core and intermediate zones, and on the border between the intermediate zone and the outermost edge of the Mediterranean world-economy (fig. 2).

The hypothesis of a geographical shift of mediating functions is compatible with all the available data and in particular with developments observed in a number of different regions. The hypothesis according to which the Mediterranean objects in the Marne-Moselle Culture were brought back by mercenaries who had fought in Italy²¹ is less convincing. because it is partial. It is also weak because most of the graves are earlier than the Celtic migrations of the early 4^th c. BC. The hypothesis assumes that large numbers of Celts went to Italy before this date, but there is no sound evidence for this. On the contrary, the Celtic invasions very logically upset the trans-Alpine exchange networks, and there is indeed a considerable decrease in the number of imported vessels in the 4th c. BC.

The Mediterranean world-economy in the La Tène period

The redistribution of economic functions precisely respects the European economic system, organised in three concentric and complementary zones. It is worth examining whether the model fits the evidence available for the subsequent development of the La Tène period. In the course of the 4th and 3rd centuries BC, Celtic groups colonise vast areas. They settle in northern Italy, within the first circle. Their direct and lasting contact with local urban civilisations leads them to adapt, and by doing so adopt not only a model of artistic expression, but probably also a form of urban organisation. The first oppida appear in the Po basin in the 3rd c. BC²², and in trans-Alpine Gaul a century later. Celtic extension into ‘barbarian’ Europe is also very pronounced (fig. 3).

It appears that the Celtic Complex, initially situated on an advantageous central arc within the intermediate zone, was expanding to completely encircle the inner Mediterranean zone which controlled the economy. In other words, the aim was to gain control over the peripheral regions most interested in trans-European exchange. However, this strategy was contradictory in the short-term: Celtic communities moved to take better advantage of exchange, but at the same time disrupted the trade networks and received fewer goods. The geographical expansion of the Celtic Complex essentially takes place within the pre-existing framework, even though this initially threatened and weakened the centre of the system and thus ultimately the system itself.

We have little information for the end of the 3rd and the early 2nd c. BC Classical texts indicate that the geographical expansion of the Celts was brought to a halt everywhere. It seems likely that socio-economic restructuring took place, preparing the ground for emergence of oppida from the late 2nd c. BC onwards (fig. 4).

The 2nd c. BC is marked by a series of Roman conquests around the western Mediterranean basin. In southern France, Roman control significantly extends along the Rhône and along the Aude up to Toulouse on the Garonne. These are in fact the main channels in western Europe between the Mediterranean and the north of the continent. By doing this, Rome had conquered almost all of the first circle of this system. The second circle becomes the zone of Celtic oppida. It is noticeable that the 5th c. BC northern limit is maintained. Some of this limit is reconfigured in the northwest, and this cannot be totally explained by our model.

During the 1st c. BC, oppida become embryonic towns. The economy of the Celtic Complex evolves towards a monetary form. Social organisation develops along early state lines. Intensification of exchange with the Roman world leads to the establishment of Roman merchants in Gaul. The socio-economic gap between the first and second circle is reduced. Symmetrically, in the third circle, rich burials containing Roman imports become more frequent²³. It seems clear that the Mediterranean world-economy is now organised around Rome. As a result of this change, the system is modified once more by the transfer northwards of economic functions.

In the 1st c. AD, the Roman Empire stretches out even further to include the second circle, which is then blended into the first. Logically, a new intermediate zone should then take shape further north, within the former third circle. The famous princely burials of the Lübsow group²⁴ are indicative of a phenomenon comparable to that of the 6th and 5th centuries BC further south. They are situated some distance from the first circle, between 200 and 600 km from the northern edge of the Empire (fig. 5). Quantitative analysis of Roman imports in Free Germania, carried out by Lotte Hedeager²⁵, supports this observation. Whereas everyday products like Samian ware and fibulae are dominant along the frontier, principally in the non-Romanised Celtic oppida zone, prestige goods – especially bronze, silver and glass vessels – are mainly distributed along the strip of land containing the princely burials.

These observations clearly deserve more detailed investigation. Nevertheless, they combine to form a mass of evidence which fits the Mediterranean world-economy model. The model provides a satisfactory match for the basic, long-term trends in European socio-economic evolution from the 9th c. BC onwards. The model does not of course account for all the details of this evolution. Additional factors operated at another spatial scale, the regional or micro-regional level. It is in fact obvious that peripherisation around the Mediterranean did not function in a uniform manner, like waves without obstacles. More intensive exchange axes are discernible along the European peninsula’s trade axes. The responses to Mediterranean demands were variable, depending on the quality and quantity of the raw materials available, but also on the socio-economic organisation which permitted the exploitation and exchange of products.

To forget the system’s interactive dimension would lead to reproduction of the old diffusionist model and its setting up as a Deus ex machina. A probably rewarding line of research is the identification of several specific systems functioning at different scales and the simulation of their interaction.

One tends to hear either openly hostile or disillusioned comments about attempts at modelling. Absolute relativism, like putting off general interpretation until a later date on the pretext of insufficient evidence, is both illusory and sterile. This results from the optical illusion created by telescoping different scales of analysis. Economists who quite rightly distinguish between macro- and micro-economy offer an example that is well worth following.

The only way forward is to build theoretical models to uncover the significance of the various factors involved. Surely one of the functions of the model is to identify departures from the rule, the investigation of which will lead to richer interpretation.