Chapter 8: Transition 5: Modelling the territorial mutations of the Iron Age in southern Gaul (6th-5th centuries BC)
Entrées d’index
Keywords : Agglomeration, Gaul, Iron age, modelling, power, settlement system; simulation, trade, transition
Remerciements
With the contributions in appendixes of Stefani Crabtree, Department of Anthropology, Washington State University (Pullman), and MSHE C.N. Ledoux USR 3124, CNRS / Univ. Bourgogne Franche-Comté, and Jacques Ferber, LIRMM, CNRS/Université de Montpellier
Texte intégral
1The fifth transition studied in the framework of the Transmondyn programme deals with a relatively short period compared to the preceding ones – on the order of several centuries – and with a limited territory, covering the coastal zones and the immediate hinterland of the Mediterranean arc around the Golf du Lion (figure 1). This transition focused on the mutations which affected the settlement system between the extreme end of the Bronze Age (around 1000-900 BC) and the beginnings of the second Iron Age (in the course of the fourth to second centuries BC).
2We present this transition on the basis of the archaeological data available, derived essentially from the Languedoc region. After delineating with broad strokes, the historical context of the transition, we will paint a synthetic picture of it, following the dimensions established in the framework of the research project (cf. chapter 2), relying principally on the synthetic studies of Michel Py1 2and Dominique Garcia3. An initial formalisation of the processes at work in the transition has been realised with the help of box-and-arrow diagrams. Finally, in the form of appendixes, we give the preliminary results of the first two attempts at agent-based modelling performed on several aspects of the transition.
Figure 1: Map of locations of the archaeological sites mentioned

The historical context of the transition
An economic and social revolution
3At the beginning of the first millennium BC, the human communities that occupied the space in question lived in a framework largely inherited from the epochs of recent prehistory (second millennium). If one witnesses an initial grouping together of the population in agglomerations often situated in elevated locations, of small size, temporary – the hypothesis of seasonal mobility of the herd-migration type between the coastal fringe and the hinterland is still debated – settlement in the main was not yet concentrated. The subsistence territory constituted the ordinary spatial framework of the groups – largely undifferentiated from one another from all points of view – where they engaged in productive activities essentially oriented towards consumption and almost exclusively developed in a domestic and/or collective context. It appears from the available documentation that the social organisation at that time corresponded to the ‘acephalous tribal’ type 45characterised by domestic production, ‘c’est-à-dire un système où chaque maisonnée a un comportement visant uniquement à la satisfaction de ses propres besoins […] sans aucune division du travail autre que par sexe au sein de chaque unité’ (that is, a system in which each household behaves in a way aimed exclusively at satisfying its own needs […] without any division of labour within each unit except according to sex)6. Because self-sufficiency is the norm regulating the means of production, exchanges are marginal and limited in quantity. When existing, they are principally of a symmetrical nature between communities. It is only from the seventh century that allochthonous objects appear and only on a few sites, revealing contacts with distant regions – continental Greece, greater Greece and the Italiot world 7. Adopting a single narrow perspective on this over-hasty panorama would be absurd: all of this is true on the average for a large part of the population of this period, but some exceptions to the rule show that evolutionary stirrings, which later periods would augment and systematise, were already active in the society’s economic and political structures.
4The geopolitical landscape of southern Gaul was abruptly transformed around 600 BC by a major historical event: the Greek (more precisely Phocaean) colonial foundation of Massalia (Marseille). This foundation was very slightly earlier than that of Ampurias on the north-Iberian coast and was soon followed by that of sub-colonies all along the shoreline: Antibes, Agde, Olbia, then Nice … The kind of colonisation here based on the ‘trading post’, is devoted to the emporía, which ‘est réellement l’alternative développée par les Phocéens en réponse à la pauvreté de leur territoire’ (is really the alternative developed by the Phocaeans in response to the poverty of their territory)8, and which is opposed in its nature to the settlement colonisation which would later be practised by the Romans, especially in the same spaces (cf. chapter 9). The territory directly controlled by each colonial establishment (chôra) along the shoreline was narrow – several km2 at most – while the hinterland remained, strictly speaking, under indigenous domination.
5These processes were not uniform, with chronological and geographic discrepancies that is, according to specific rhythms and modalities here and there. Nevertheless, the indigenous world experienced in all domains, over the course of the sixth to fifth centuries BC, profound mutations which distanced it, little by little but more and more radically, from the state which we have just rapidly described. The issue remains of knowing what part was played in the processes of social transformation by the intrusion of the colonial phenomenon on the local scene. In other words, one ought to measure what in the evolution of native societies is due to endogenous processes, supposed to unfold more or less in the same manner without the external context which imposed itself on them, and what is more rightly attributable to foreign contributions. And, a subordinate question: should one consider that these external agents functioned as simple stimuli in an evolutive chain otherwise autonomous, or, rather, did they provoke a true bifurcation in the indigenous fundamentals?
6At the risk of being accused of reductionism through schematisation, one can summarise in broad strokes the principal characteristics of the general process which led from the initial state of society during the transition from the Bronze Age/Iron Age to that which is observed five centuries later. The essential vector of economic mutation in the first half of the first millennium BC, at the origin of all the other transformations of the indigenous society, was the progressive increase of surplus production. The mechanisms of this phenomenon’s appearance and development were no doubt multiple and complex: natural cyclic tendencies, internal development of indigenous productive forces, clearing of new land for farming, agro-pastoral technical process, climatic conditions9 …
7The process of producing increasing tradable surpluses, once underway – perhaps sustained and reinforced by external demand – entailed the complete transformation of the social organisation, in accordance with one principal tendency, which was towards greater complexity and hierarchisation. Indeed, the necessity always to produce more and of better quality so to create exchange value rendered structurally unsuitable the traditional frameworks of the domestic economy, as defined above. All the vectors of increase in production (improvement of yields, adoption of new techniques, specialisation of production) imply the disappearance of the single polyvalent actor of the ancient system – at once shepherd, farmer, fabricator of all the material possessions he used every day – and fostered the segmentation of society into classes through specialisation of functions. Thus, alongside a very large majority of the population which would remain attached to primary production – direct collection and exploitation of resources – there appeared a secondary sector of specialised artisans engaged in activities of transformation. Then, almost simultaneously, a collection of tertiary sector superstructures was created, soon brought into being by the growing complexity of the social network: functions of commerce and trade, functions of power, military and defence functions, religious functions, etc.
8It cannot be imagined that a system of this sort could function within the framework of a diffuse distribution of small settlements on the territory. To be effective and correctly join together all these functions almost necessarily implies the physical grouping of inhabitants in agglomerated poles matched with a territory: this was precisely the role that would be played by the oppidum-cities which became the norm in the entire Mediterranean region beginning with the second Iron Age.
The mutation of the settlement system
9Among all the phenomena observable during the period, our main interest concerns the mutation of the settlement system over the course of Protohistory. The transition during the sixth and fifth centuries was principally characterised, from this point of view, by an increase in the number of settlement sites, supported by a concentration of the population. This concentration was reflected in a process of urbanisation and a phenomenon of social differentiation - such as M. Py et D. Garcia have described - and in accordance with a gradient extending from the coast to the hinterland10 11. Whatever the exact period and the extent of the phenomenon of population clustering, it was generally associated with development of the trade-economy, which is attested, thanks to excavation or prospection, by the presence of imported objects. At the same time, the multiplication of cereal storage structures is partial testimony to the increase of production surpluses potentially for trade purposes, even if cereals undoubtedly do not constitute the only possible counter-value. The specialisation of activities is both the cause and the consequence of the increase of production surpluses. From the social point of view, this specialisation of activities engenders a specialisation of the roles of individuals in indigenous communities, indeed of the communities themselves, which may even develop systems of cooperation and alliance or even, on the contrary, a competitive dynamic.
Description of the transition of the settlement system between 900 and 200 BC
10Following the principle established for all the transitions (cf. chapter 2), we have divided the period into two regimes separated by the transition we are dealing with. These two regimes are described according to the thematic fields selected in the framework of the research programme (see figure 4 of chapter 2): way of living/use of space, movement/transport/type of mobility, ways of feeding/relationship to the environment, social organisation/form of social interaction, power type/administration/territorial mesh. The first regime corresponds to the end of the Bronze Age and the first ancient Iron age, that is, to the period roughly from 900-625 BC. In certain respects, this period might in itself contain a transition, for according to D. Garcia, the forms of habitat seem to evolve in a rather pronounced fashion over the course of these three centuries. However, since the material culture, exclusive of importation, remains relatively unchanged during the whole period, we consider it as involving a homogenous regime. By contrast, from our point of view, the transition is marked by the development of regular exchanges with the Mediterranean ‘merchants’, especially Etruscan and Greek. The mutations of the settlement system appear at the very end of the seventh century and the beginning of the sixth, but become decisive only over the course of the transition that concerns us here, between 625 and 400 BC, with a massive phenomenon of urbanisation and the development of a dense network of habitats. The second regime, between 400 and 200 BC, corresponds to the period of development of the agglomerations of the second Iron Age within a socio-political system of the ‘chefferie’ (chiefdom) type12 and constitutes the beginning of the first regime described in transition 6 (cf. chapter 9).
11For these periods without written sources, it is important to emphasise that the hypotheses proposed regarding the political and social, administrative and territorial functioning of the society rest entirely on the interpretation of archaeological data in the light of anthropological theories most often founded on references derived from modern societies in radically different contexts. The proposals formulated for each transition regime are therefore deliberately schematic, debatable and debated. However, we have accepted them as frameworks for reflection with regard to what follows.
From one regime to the other, from ‘fragmented’ to ‘continuous’ territory
12For the first regime, the “way of living/use of space” is defined globally by a great variety of types of habitat, relatively small (between 0.2 and 2 ha), most often dispersed or sometimes grouped in proto-agglomerations of several hectares13, characterised by an irregular, unplanned spatial organisation and an architecture in perishable materials. By contrast with the agglomerations of the second Iron Age14, these types of habitat are rarely fortified, but certain may be enclosed within natural defences, ditches or raised earthworks, perhaps surmounted by a palisade15. During the second regime, agglomerations become by far the most common type of habitat, while the dispersed habitat regresses greatly, especially in the third century, with the abandonment of a large number of sites. The growth in overall surface area occupied in Languedoc16, the increase in the number of domestic units in the grouped habitats – in particular in Provence17 –, the increase in the size of the houses, now with multiple rooms, testify to a real demographic dynamism. The communities occupy the space in a more or less loose and discontinuous fashion during the first regime; during the second, by contrast, a much more regular pattern of settlement structures the territory, with normalised spacing between the agglomerations, respecting their functional level.18 19.
13Despite a well attested sedentary way of life during the first regime20, the mobility of all or part of the communities in search of additional lands to ensure their subsistence production is not excluded21. In this perspective, the hypothesis of a territory rather ‘fragmented’22, accepted with nuances by a number of scholars, implies a mode of occupying the land which would not be centred principally on the habitat but rather on the axes of displacements, the locations of activities (production and exchange) and also on the necropolises, to gather the dead of the community, all this within a regional space limited overall23 24. According to M. Py, metal-working, which still remains exceptional, seems also to follow the same pattern, with the probable existence of itinerant artisans producing objects of decoration and grooming, tools and weapons, on demand.
14In the second regime, the increase of production surpluses, which in theory leads to a proportional storage capacity, should encourage regular trading. However, M. Py stresses that the archaeological evidence points instead to a diminution of storage capacities over the course of the period, which goes along with a drop in the quantity of imported objects. This phenomenon might be interpreted as the sign of a crisis, but this would be wrong because according to the indicators of the ‘way of living/use of space’, this period instead emerges as a phase of growth. This is why one should undoubtedly postulate that the reduction of storage volume observed is paradoxically due to a more rapid distribution of the production in a better organised trading system. The concomitant diminution of the volume of imports might be evidence of an investment in other forms of social enhancement (fortification, for example) when a relative indifference of the local communities is developing toward the exotic products considered as prestigious during the first regime 25. The augmentation of productive capacities is also reflected in the development of a true craft sector (iron and ceramics) as well as the appearance of a system of calibration for the trade in Mediterranean objects with the emergence of the first evidences of money (even if its usage and hoarding remained very modest)26.
15The two types of spatial occupation mentioned above are explained by the ‘ways of feeding/relationship to the environment’ and the relationship between the communities and their ‘environment’. Under the first regime, an economic system of subsistence founded on agro-pastoralism, supplemented by a fishing and hunting27 activity compatible with the frequentation of temporary locales, is dominant. During the second regime a system of production is developing that, although still essentially domestic in its frameworks, now creates increasing surpluses of agricultural production which imply, de facto, a more stable structuring of the community’s land. Thus, for example, the increase in the proportion of bovines – used as draught animals – among the livestock reveals new farming practices (keeping animals in stalls), which imposes fixed dwelling for the populations. The durable appropriation of a section of land, or at least of its productive capacities28, and the introduction of species providing deferred yields (vines, olive trees)29 30as well as regular improvement of the soil31, constitute further signs of the economic and territorial transformation of the second regime. Nevertheless, several large sites, around which a diversified productive area began to be managed and tended to become stabilised, stand out in the overall pattern from the end of the first regime and look forward to the transition32 33.
16On the level of ‘social organisation/form of social interaction’, the optimisation of the production and of its distribution engenders a specialisation of individuals in craft-production and commerce. According to M. Py and D. Garcia, the workshops discovered in excavations multiply (potters, forges) from the end of the fifth century, and inscriptions – on strips of lead – suggest the existence of indigenous “brokers” at least from the fourth century. While the society of the first regime undoubtedly rests on a fairly egalitarian base, one simultaneously perceives, in the specialisation of productive functions, signs of the emergence of a more hierarchised social structure. Thus a warrior caste reveals its presence from the beginning of the transition, indeed from the extreme end of the first regime,34 35by way of the sculpted statuary of men with weapons and the public exposition of sacrificed severed heads and weapons symbolising the spoils of vanquished enemies. This governing class also calls attention to itself by the development of prestigious constructions (figure 2) and by the deposition in its tombs of arms, horse-bits and military equipment36. In the more common necropolises of the population, funerary offerings were relatively rare and standardised until the seventh century, as would seem indicative of a social levelling. It is only beginning in the fifth century that the appearance of isolated tombs seems to manifest a social hierarchisation reflected in different class levels and different statuses among members of the community.
Figure 2: Towers and curtain wall of the oppidum of Nages (Gard), third-second centuries BC

17The evolution of the social functioning testifies to the emergence of a power structure which archaeology by itself is insufficient to reveal. The existence of this power structure becomes evident when one looks at the relationship between communities and the territory over time. In the first regime, the pattern is founded on a group of complementary land parcels and not on the appropriation of a continuous space. Therefore the relation of the communities to the territory is rather fluid, changing according to the perspective or depending on the activities, seasons... The appropriation and perception of a continuous territory will become characteristic of the communities of the second regime. During this regime, the way the territory is appropriated manifests the desire of certain individuals to set themselves apart from the rest of the group but also, and in particular, to situate themselves within the geographic space in a singular fashion related to others communities: this is an expression of territory-marking. In the same epoch, the identification of cultural groups – based on studies of their material culture – associated with observation of the regular distribution of fortified agglomerations (oppida) supports the idea of a division of the space into territorial units undoubtedly forming a hierarchised network, whose levels, although perceptible, are complex and difficult to interpret from the point of view of political organisation.
Schematisation of the transition and description of several processes
18After describing the transition, the procedure consisted in formalising the two regimes concerned, called A and B, using arrow diagrams, concentrating on the diverse processes which, from our perspective, underlie the transition.
19In the first regime (A, figure 3), the seasonal mobility of part or all of the group undoubtedly favoured a frequentation of varied ecosystems with complementary resources based on dispersed dwelling places and movement, while a portion of the fixed settlement may have supported the agricultural development which permitted the increase of surpluses (A-7). Such agricultural surpluses (A-137) allowed the group to subsist with a minimum of storage38. This encouraged the progressive diminution of hunting activities for obtaining food (A-2) and also enabled creation of the counterpart necessary for trading (A-3) with other local or exogenous communities (Etruscan and Greek, for example).
20The process of releasing tradable surpluses, once set in motion – perhaps supported by external demand – entered into the positive feedback loop. Contacts with ‘exogenous’ actors and trading enabled the emergence of a social hierarchisation (of the ‘proto-chief’ or ‘Big Man’ leader type39), which in turn, it may be hypothesised, stimulated the organisation of production and trading (A-8) and (A-9). This question is the subject of debate among scholars: why did the surpluses increase? Was this the result of demography? Or of the stimulation provided by trade? What were the respective roles played in the process by the internal and external actors? And from what point did the phenomenon become irreversible?
Figure 3: Arrow diagrams of regimes A and B of the transition

21Whatever factor may be responsible, this mechanism, linked to trading by successive reinforcements, would entail the complete transformation of the social organisation, according to a primary tendency to complexification and hierarchisation whose affirmation is perceived in the second regime (B). In this regime (figure 3), the production of surpluses is still active (B-1) and enables the reinforcement of trade (B-3). In turn, production is stimulated by trading, but what is chiefly noteworthy is that it becomes diversified and specialised, especially with the emergence of viticulture (B-2). In addition, the development of more intensive agrarian techniques (using manuring) encouraged an increasing agricultural productivity and, potentially, the amount of the surplus (B-5). In our system the goods figuring as exchanges for Mediterranean products are reduced to the category ‘craft production’, which covers all non-agricultural products (minerals, metals, ceramics, textiles). In this diagram, only the trading of slaves, whose existence is highly probable, has not been taken into consideration.
22The emergence of an elite in regime A, which is slightly apparent in the necropolises (A-10), becomes more prominent in regime B by virtue of the greater number and quality of offerings (B-9). The recent chronological reinterpretation of protohistoric statuary gives reason to assume the early manifestation40 of a social elite in some sanctuaries. These sanctuaries undoubtedly had the function of intracommunity spaces41 and might have served the dispersed populations as assembly places for exchanging goods (A-11). In the second regime, hierarchisation is chiefly evident in the assertion of control over those spaces (B-10) on which agglomerations of the ‘oppidum’ type are established (B-11). The integration of the sanctuary within the settlement, associated with the unprecedented development of fortifications, is a sign of the demonstration of power and of territorial control (B-11).
23In regime A, the progressive increase of surplus has enabled the development of various activities, such as craft production, which in turn have improved agricultural productivity (fabrication of iron tools) and potentially led to additional surplus (A-6). The emergence of a category of itinerant craftspeople, in a sedentary system but one in which a number of individuals were mobile on a seasonal basis (A-4), was also able to foster exchanges between the groups (A-5). The necessity of always producing more and of better quality to create exchange-value implied the disappearance of the polyvalent actor of the ancient system and encouraged the emergence of new classes by way of the specialisation of functions. The development of the grouped and planned settlement (oppidum), and in particular of indigenous trading posts, favoured this emergence of individuals practising a specialised activity (B-13): crafts-people, merchants. In turn, we may form the hypothesis that the development of social hierarchisation had an influence, on the one hand, on the modes of production (B-8), on the other hand, on the nature of contacts with the ‘exogenous’ actors and on the modalities of exchanges (B-7). Without a doubt, at this level there was a feedback loop between the exchanges and the transformations of the mode of production that stimulated the pyramidal structuring of society (chiefdom, warrior aristocracy). This actually gave local actors the possibility of choosing and negotiating over goods of Mediterranean origin, particularly from the point when the indigenous shoreline trading posts were well established. In this context, the question arises of the installation of Mediterranean merchants from the end of regime A and the development in regime B of colonial trading posts or colonies, on the model of Marseille (B-12).
24It is thus that in regime B, alongside a portion of the population, of which by far the majority remained attached to primary production, there appeared a secondary sector of well-established artisans, then, virtually at the same time, a group of tertiary sector functions (commercial, political, military) in response to the increasing complexity of the social network. To be efficient and properly joined together, all these functions imply, almost of necessity, their physical grouping in the agglomerated poles surrounded by a territory (B-6): this is precisely the role which the oppida-cities would play, connected by a durable network of communications (roads, ports) (B-14), which constituted the norm for human settlement in all the Mediterranean region beginning with the second Iron Age. The dispersed habitat then became marginal at best, while the small settlements were established in the areas of territorial influence of the oppida.
Discussion
25The merit of this work is to offer a model of the systemic type (arrow diagrams), taking into account a considerable sum of archaeological knowledge which has already been the object of a number of synthetic publications. From this rich but necessarily complex picture, based on hypotheses deliberately selected and assumed for the exercise, our objective has been to separate analytically the processes of the evolution of the settlement system linked to the stimulus of long-distance trading. The point of view defended here regarding the period under consideration is therefore highly partial and schematic. However, through it the explicit characterisation of a number of mechanisms has been possible. These mechanisms have contributed to complexifying the settlement system because, if the settlement system mutations can be understood by the observation of different archaeological data, those data alone are not necessarily able to explain changes.
26Despite our efforts to simplify the transition observed, it has not been possible to propose a simulation model founded on the entire formalisation developed here. The difficulties with which we have been confronted are of several kinds. The first was linked to the complexity of the transition, which, not surprisingly, implies a number of imbricated processes leading the settlement system to undergo such a transformation that it opens onto a new social and territorial conception. The second obstacle was paradoxically due to the richness of the modelling tools, which led the researchers interested by our transition to develop modelling experiments totally independent of each other. In addition, the experiments carried out within the framework of Master and PhD academic work involved their own constraints. Thus, two drafts of models were developed in autonomous fashion and in two different environments. Both are based on the driving force of trading exchanges but concentrate on two typical phenomena. The first model developed in Netlogo focuses on the competition between imported goods, as perceived in the material traces, and the capacity of the exogenous actors to keep settled in the territory (annexe 8.1). By contrast, the second model focuses on the impact of trade on the emergence of specialisation in the native population. The aim is to simulate the phenomenon of hierarchisation, which, in part, characterises the transition. Moreover, this last experiment had the objective of testing and improving the environment of MetaCiv multi-agent modelling presented in appendix 8.2.
In regime B (figure 9), the exogenous exchanges involving wine create the particular value of this new product (appearance of the culturon ‘value of wine’), considered as socially valued. This entails development of agricultural production (cereals) by the farmers, who need additional reserves (surplus) to satisfy this trade and invest in tools in order to be more efficient. Such behaviour supports the culturon ‘demand for tools’. The culturons ‘reserves’ and ‘value of tools’ are thus progressively emphasised (figures 7 et 9), and this consolidates the specialisation of the classes of agents ‘farmers’ and ‘artisans’. Finally, this process stimulates profitability and the opportunity for agents to invest in other kinds of socially valued goods, something which leads to the emergence of the cogniton ‘merchant’ in the minds of indigenous agents. In our model, the agents of exogenous population ‘Mediterranean merchants’ are indirectly integrated through the object ‘wine’. The growing synergy of local and exogenous exchanges triggers the culturon ‘prestige’ in the indigenous population, and this is at the origin of the social hierarchisation figuring in regime B (cf. chapter 8).
Annexe
Annexes
Appendix 8.1. The simulation of trading in the first Iron Age: a highly stylised model
By Stefani Crabtree.
The transition, as described in this chapter, shows the establishment of a complex relation between the local actors and the Mediterranean merchants. A simplistic view would consist in advancing the hypothesis of a local system progressively constrained by the colonial power with an unequal balance in the exchanges, forcing the local actors to produce more and more surplus as a counterpart to the imported products. Given the manifest signs of prosperity of the communities in the second regime (B), such a hypothesis requires discussion in order to explore in greater detail the role of local actors within the mutation of the system from the first regime (A). The point of view adopted here is an attempt to understand the process underlying such a radical change in the trade in Mediterranean products42. Considering the archaeological evidences observed within the indigenous settlements, it has been demonstrated that the amphorae of Marseille (the Phocaean colony of Massalia) supplanted the Etruscan and Greek productions imported from the end of the sixth century (figure 4)43. Does this reflect a rapid modification of the terms and nature of exchanges when the Greek colony of Marseille obtained a regional monopoly at the expense of the Etruscan merchants44 ? Or is this change due to a change in the behaviour of the indigenous populations, affirming their preference for the Massalian products to the detriment of the Etruscan ones, and this in the context of a redefinition of the networks of relations among locals, Etruscans and Greeks? If a definitive answer is out of reach, the role of the local actors in the long-distance trade network can be examined in detail to assess their potential contribution to the process of substituting one source of production for another. In any case, the indigenous communities were undoubtedly involved in a competition linked to trade. This competition found an expression locally with regard to habitat (grouping, ostentatious fortifications) and through funeral practices (social differentiation).
Figure 4: Evolution of the proportions of amphorae by category in the region of Nîmes

1) Etruscan, 2) ancient Greek, 3) Greek from Marseille (Massalian), 4) Italic.
(Source: after Py 1990)
3.1.1 Description of the model
Using Netlogo45, a highly stylised model was conceptualised and developed whose the general features and the principal results obtained are exposed here. The model aimed to determine : 1) whether the Mediterranean merchants could survive (hence, become established over the long term) without being farmers, counting exclusively on the grain trade for food; 2) whether the pattern observed in the archaeological data may emerge from a series of exchanges between ‘individual’ agents (locals (Gauls) and Mediterranean merchants), without any strategy of overall competition or cooperation on the scale of the communities studied (indigenous and Mediterranean). The simulation is based on two mechanisms: on the one hand, a simple extraction of resources in the environment; on the other, a system of exchange between two types of agents. In order to confront the results of the simulation with the pattern observed in the archaeological data (figure 4), a time step of one year with 500 iterations was adopted. The output of the model is of two kinds: 1) a population curve (number of agents); 2) a curve of the volume of amphorae traded. In the model, the autonomy of the individuals is deliberately enhanced, as was obviously not the case in reality, since the consumption of wine was above all reserved for the elites. On the theoretical level, in a virtual world, it is a matter of seeing whether the simple modification of some parameters can provoke a transition perceptible through the number and nature of the kinds of amphorae traded.
The space in which the agents evolve is a grid of 80 by 80 cells. The area is composed of a portion of sea with a band of shoreline and a portion of land potentially usable for producing cereals or grapes for wine, that is, goods for trade. The cells corresponding to lands represent a surface of 5 hectares, i.e. the average size of a farm. Despite the wide range of goods exchanged (salt, fish in brine, cured meats, oil, resin, iron ore, even slaves)46, as is well attested by archaeological and historical data, in the model the products traded are reduced, to cereals, measured in hectolitres, in return for wine, measured by the number of amphorae. The rate of consumption necessary for the survival of the agents and the rate of exchange are fixed by these two types of goods. The agents are of two types: on the one hand, local cereal producers (‘Gauls’), on the other hand, (Mediterranean) ‘merchants’. The merchants could settle and produce wine (amphorae) locally, but they had need of cereals to survive. It is important to note that an agent represents in reality one unit of family production. In the model, the unit of family production may be equated with one single individual or several (according to the level of stock available). One unit of production may be divided into two, with one moving in order to settle elsewhere. The two new units then share the stock initially available. A minimum of 10 hl of grain is necessary for the survival of one unit with one individual, that is, an average annual consumption of 6 hl per individual47, to which must be added 4 hl for the cereals needed for seeding and for the energy needed for harvesting48. As a hypothesis, the model considers that a unit of production will be able to cultivate from 5 to 15 ha (more in the case of the largest families), that is, from 1 to 3 cells.
3.1.2 Initial conditions
At the beginning of the simulation, iteration 0 corresponds roughly to the date of 700 B. C., the Gaulish agents (150) are distributed randomly in the (land) space, and they are endowed at their ‘birth’ (creation) with 20 hl of stock. One hundred merchant agents are initially created, and they possess a stock of 60 hl of grain and 20 amphorae. The arrival of a first wave of (Etruscan) merchants intervenes after thirty iterations, and the second wave of (Greek) merchants after a hundred iterations.
Each agent has a lifespan of a maximum of 80 iterations, with a probability of decease of 50% at each iteration and a probability of reproduction simulated through the division of a unit of production (3 % per iteration).
The resource production by cell and by iteration is 20 hl of cereal for the Gauls and 10 amphorae for the merchants. The annual consumption of cereals is fixed at 6 hl per agent, given that 4hl are also necessary for seeding and harvest. The rate of wine consumption is set at 1 amphora per year for each agent, but a stock control parameter also includes the loss of one amphora per year.
Trading is regulated by a 5% probability of selling wine per iteration for the merchants and 1% probability of purchasing for the Gaulish agents. The rate of trade of cereals for wine is a model parameter that varies from 20 to 60 hectolitres of cereals for 10 amphorae.
3.1.3 The main outputs of the model
In an initial step, the setting of the parameters of the model aimed at testing the hypothesis that the development of the Etruscan trade led to an intensification of agriculture in the space under study, thereby creating additional surplus and being able to respond to the trade in Mediterranean goods49 50. The model was then calibrated by defining the parameters able to 1) sustain the population of ‘Etruscan merchant’ agents, depending on their ability to trade; and 2) assure the survival of the local farmers. To maintain the balance, an optimal exchange ratio of 40 hl of grain for 10 amphorae (40/10) was chosen. The simulations show that under the ratio of 30/10, the population of merchant agents disappears very quickly and that over 60/10, it is the local Gaulish producer agents who perish rapidly.
In a second step, two types of merchant population were considered to measure the effect produced on the distribution of kinds of amphorae at the time (figure 4) and to test the capacity of each agent to survive. In this new model, a parameter of ‘preference’ was introduced. This makes it possible to define the probability that a Gaulish agent will chose an Etruscan wine or a Greek wine (from Marseille) as a function of the value (taste) he assigns to each51. Each combination of parameters with three distinct values of preference has been repeated 30 times, i.e. 330 simulations in all, to provide average curves (figure 5). In the end, the criterion of ‘preference’ has an influence on the results obtained but not enough to explain the evolving accumulations of amphorae during the period. A preference of at least 30 % in favour of the new merchants (Greeks from Marseilles) is necessary for them to survive and establish themselves in the system of exchange with the Gaulish communities (figure 5 a et b). The preference of the Gaulish agents for the Massalian amphorae must attain at least 80 % to get more amphorae from Marseille than Etruscan amphorae (figure 5 c to d). It is only with a monopoly of preference for the Massalian amphorae (100 %) that the Etruscan amphorae disappear definitively (figure 5 e et f).
Taking the time step of one year as a basis, it should be noted that, in any case, the process of replacing one type of amphora by another is much slower in the simulation than in reality. This would lead one to suppose that individual preference did not by itself account for the relatively abrupt change observed in the archaeological data (figure 4). Nevertheless, the model deserves to be refined and experimented with further before validating such a hypothesis. Indeed, it is important to stress that the model does not take into account the potential capacity of the Etruscan merchants to cooperate with the Greek ones from Marseilles to trade part of their merchandise.
In conclusion, considering the role played by the Gauls in the evolution of the trading system, this highly stylised model shows the intensity with which they must have acted to provoke the change observed in the archaeological data. The simulation outputs suggest, however, that the choices of the agents alone are insufficient to explain this. This fact invites much more serious consideration of the hypothesis of politico-economic cooperation between Etruscan and Massalian merchants52 .
Figure 5: Evolution of the population

(number of ‘Gaulish’ agents in red, ‘Etruscan merchants’ in blue and ‘Greek merchants from Massalia’ in rose) and of the type of wine chosen (total number of Etruscan amphorae in blue, and Greek from Massalia in red) as a function of the individual preference of the agents.
The level of preference in % is indicated in the centre (cf. note 56). The solid coloured line indicate the average value, while the dotted coloured lines correspond to a standard deviation above and below the average. The lines in grey record the global variation of all the combinations of parameters.
Appendix 8.2. MetaCiv, a meta-language for modelling an aspect of the transition: the emergence of specialisation within the settlement system
By Jacques Ferber, Thérèse Libourel, Laure Nuninger, Julien Nigon and Gauthier Maillé.
Through the model presented in this chapter an experiment was carried out with a generic framework of MAS (multi-agent system) furnished with a meta-language - MetaCiv. MetaCiv facilitates the formalisation of complex social systems and has the advantage of being richer on the conceptual level. While the Netlogo model (appendix 8.1) depends on a monolithic individual-centred approach (the behaviour of the agents being fixed in the simulation), MetaCiv takes into account the evolution of the behaviour of the agents on both the collective and the individual levels.
In performing this experiment, our objective was two-fold: on the one hand, the idea was to formalise an aspect of the transition, and, on the other hand, using this example, to clarify the role of the various meta-language elements, as well as their operational capacity within the software prototype, which, in turn, has been substantially improved.
The MetaCiv platform: theoretical foundations and principles
MetaCiv is based on the meta-model MASQ (Multi-Agent Systems based on Quadrants)53. Starting from a general conceptual framework, MASQ uses an approach by quadrants54 55derived from the work of Wilber56.
The generic architecture of MASQ is based on four quadrants delimited by two axes. The first axis contrasts the individual with the collective aspects (see fig. 6), while the second axis contrasts the interior with the exterior points of view. The Individual-Interiority quadrant (I-I) concerns the mind, the mental state of the individual agents (their beliefs, desires, intentions). The Individual-Exteriority quadrant (I-E) describes the physical bodies of the individuals, their behaviour, the objects used. The Collective-Interiority quadrant (C-I) concerns the collective knowledge and beliefs shared by groups of agents, while the Collective-Exteriority quadrant (C-E) concerns social structures and organisations.
Figure 6: MetaCiv meta-model describing the elements of the generic architecture of the agents within the quadrants.

In MetaCiv, each agent has a physical body (quadrant I-E) separated from its mind (quadrant I-I). In addition, the agent’s architecture is composed of two fundamental elements: ‘cognitons’ and ‘plans’.
Cognitons are ‘elementary cognitive units’ that may be dynamically added to or withdrawn from the mind of the agent. At a given moment, they constitute the dimensions of the mental state of an agent, integrating reactive and cognitive aspects (beliefs, skills or perceptions) which enable it to take a decision.
The plans represent the different activities that an agent may perform. Each plan is comprised of a series of actions, more or less complex. A weight is linked to each plan. This weight is determined by the weights of the cognitons present in the mind of the agent and those of the links of influences. The cognitons guide decision making, and the links of influence determine the positive or negative value of a cogniton with respect to the plan envisaged. Take, for example the case of two cognitons influencing the same plan: if the first has a weight of 2 and its influence link is at +4, and if the second has a weight of 3 and its influence link is at -1, the weight of the plan is equal to 5, that is (2x4) + (3x-1). This weight can be reinforced or diminished as a function of the activities of the agents: actions. The actions can, by feedback, change the mind of the agent by reinforcement (+) or weakening (-), by acting directly on the weights of the cognitons and therefore on the weight associated with the plans.
Figure 7: Mechanism of weighting plans of action and feedbacks on the weight of the ‘cognitons’ and ‘culturons’.

The agents evolve in an environment structured by a collection of cells grouped in zones (agregates). When moving within the environment, the body of an agent is able to extract information by perceiving the environment’ properties (nature of the terrain, dynamic57, types and quantities of resources available in this zone, traces left by other agents while moving…).
The agents can use what are called objects, which modify their physical or cognitive characteristics. These objects may be of a number of types, and are generated by the actions of the agents. They can be collected, as is the case with raw materials (especially cereals, wood, minerals), or manufactured. The possession of an object, its use or its destruction is likely to modify the agent’s cognitons (a tool promotes certain qualities), as well as its physical characteristics (increase or loss of energy).
During the simulation, groups (quadrant C-E) and ‘culturons’ appear. They will enable the cultural aspects (norms, collective beliefs) to be combined at the level of the decision-making of each agent (quadrant C-I). Each role (or function) within a group carries a collection of culturons, which represent common cultural elements shared by all agents performing that role. The culturons function in a similar way to the cognitons: when an agent assumes a role, it assimilates all its associated culturons, which will then act (like cognitons) on the weights of the plans.
An application of MetaCiv on the basis of regimes A and B of transition 5
The modelling experiment aimed at testing the driving force of trade development into the specialisation of indigenous individuals. This experiment focused on a process of social construction in which the environmental conditions are extremely stylised and the increase of agricultural surplus is seen as an indicator of the degree of social organisation. Such surplus would contribute to the specialisation of the functions or roles of individuals. This specialisation allowed them to free themselves from the domestic framework, involving them in the structuring of the settlement system with the development of agglomerations: oppida (cf. chapter 8). The quantity of surplus produced may evolve gradually or abruptly, and the crossing of a critical threshold (cf. chapter 1) is necessary for the stock to be sufficient to collectively allow the sustaining of persons specialised in functions other than food production. Through the way in which individuals may evolve on the cognitive level as a function of their environment, and especially their social environment, these are the conditions of emergence of that specialisation of the population which we seek to model.
After mastering the elements of the meta-language, we attempted to re-transcribe an aspect of the expected behaviour of the actors. To this end, we created two draft models for regimes A and B in conformity with the meta-model of figure 6. Each of the models includes agents, elements of a group called ‘population’. In figures 7 and 8, the agents are endowed with a mind, that is, with a collection of cognitons (for example, ‘farmer’ or ‘artisan’) and culturons (for example, ‘need for food’), and this enables them to establish plans, or a collection of actions (for example, ‘practise farming’ or ‘become artisans’). They are also endowed with a collection of characteristics (body) and can manipulate one or more objects (inventory, here a collection of tools, for example), move, interact with their environment. In the model, the cognitons and the culturons will influence the plans in a positive or negative manner, depending on the importance the agent attributes to them (weight).
For regime A (figure 8), we focused on the way in which the agents (indigenous Gauls) reinforce their cognitons in order to specialise progressively in one or the other activity (farmers or artisans) through interactions between agents and/or with the environment.
Figure 8: Formalisation of regime A in MetaCiv, from the point of view of the indigenous Gaulish communities.

At the outset, the agents have the multiple functions of food-gatherer58, farmer and/or artisan. Thus, in the initial conditions of the model, each agent benefits from the cognitons ‘gatherer’, ‘farmer’ and ‘artisan’, with an original greater or lesser probability of each of the functions.
In the model, hunger is a driving force; hence, obtaining food resources is an important motivation which leads human agents to decide to gather food, to practice farming or exchange goods (tools for an artisan) for food, something which presupposes the release of a minimal surplus. The culturons ‘hunger’, ‘need for food’ and ‘reserves’ are shared by the whole indigenous population. In this view, only one single type of resource is involved (gathering and farming provide the same resource of food), and there also exists only one single type of tool (named ‘tool’).
In the model, the reinforcement of each of the cognitons, that is, the mechanism of specialisation, is linked to the success of the actions set in motion by a plan. For example, in figure 8, if the cogniton ‘farmer’ is activated to trigger the plan ‘practise farming’, and if the agent has a field, he can sow and harvest, a fact which strengthens his cogniton ‘farmer’. Similarly, if he possesses ‘tools’, that will increase his production and strengthen still further his cogniton ‘farmer’. By contrast, if he has no field, he will have to find one, but if his search does not succeed, he sets in motion the plan, ‘become an artisan’, something which will reinforce his cogniton ‘artisan’.
In order to reinforce his cogniton ‘farmer’, the agent may develop his productive capacity by investing in ‘tools’. To do this, he can exchange food for tools with an agent 'artisan', on the condition that he produces enough surplus. In turn, the capacity of an artisan to produce will depend on the demand for exchange of tools for food. The increase of the capacity of tool production will allow the agent to reinforce his cogniton ‘artisan’ when the action ‘create tools’ has finally succeeded. In this loop, it is therefore the surplus – that is, the portion of food reserves produced beyond the threshold of subsistence – which guarantees the process of reinforcing cognitons for each agent.
Demand for tools appears at the level of the entire population (group of agents) following the triggering of an interaction of barter. Any agent with enough food reserves (surplus) may hope to invest in tools and thus participate in exchanges (barter). The culturon ‘need for tools’ then reflects this desire; its influence modifies the weight of the plan ‘produce tools’ and thereby strengthens the cogniton ‘artisan’.
While their cognitons were initially identical, the model of regime A reveals a distinction between two groups of agents (‘farmers’ and ‘artisans’). This is due to the increase in weight of one of their cognitons: for example, an artisan may still do some farming and some gathering, but his behaviour as an artisan is more important overall. In this regime, interactions with the exogenous civilisations are minimal or reserved for the rare elites, and they are therefore not included in the model.
It is uniquely the gradual reinforcement of cognitons that favours the transition to regime B. Within this system, it is important to note that the cogniton ‘gatherer’ tends progressively to disappear. In the model the cogniton ‘gatherer’ plays only a regulating role to keep itself alive, as long as the agent has not definitively asserted its farmer or artisan behaviour.
Figure 9: Formalisation of regime B in MetaCiv, from the point of view of the indigenous Gaulish communities.

The signification of the terms, elements and links are the same as those of figure 7.
To conclude, the principal interest of MetaCiv lies in the possibility of expressing, using a meta-language (while limiting the computer coding), the agents’ abilities and plans of action. The respective influences on the agents’ decisions of different factors, internal (individual perception) and external (culture, norm, environment), vary as a function of the interactive situation, as a result of the weights calculated for each plan of action. In the model developed in Netlogo (appendix 8.1), although it was possible to test the behaviour of the agents to explain a radical change in trading observed through archaeology, that programme is relatively monolithic, compared with MetaCiv. Indeed, if one takes up the idea of the process of ‘preference’ for one good over another developed in the Netlogo model, the architecture on the basis of cognitons opens further perspectives because of the possibility of evaluating the evolution in the behaviour of agents and, retroactively, of groups (culturons). If the application of the meta-language is relatively easy and a source of fruitful interactions, its implementation here was not sufficient to pursue the modelling process to its end and produce an operational model.
Notes de bas de page
1 Py Michel, Culture, économie et société protohistoriques dans la région nîmoise, École française de Rome, 1990.
2 Py Michel, Les Gaulois du midi : De la fin de l’âge du Bronze à la Conquête romaine, Nouvelle édition revue et augmentée, Éditions Errance, 399 p, 2012.
3 Garcia Dominique, La Celtique méditerranéenne : Habitats et sociétés en Languedoc et en Provence du VIIIe au IIe siècle avant J.-C., Première édition, Paris, Éditions Errance, 207 p, 2004, Deuxième édition revue et augmentée, Paris, Éditions Errance, 247 p, 2014
4 Renfrew Colin, « Introduction: peer policy interaction and social-political change ». In Peer policy interaction and social-political change. Cambridge, Cambridge University Press, 1986, pp. 1-18.
5 Sahlins Marshall, Âge de pierre, âge d’abondance. L’économie des sociétés primitives. Paris, Gallimard, 1976, 409 pp.
6 Garcia D., La Celtique méditerranéenne 2nd ed., op.cit.
7 Garcia Dominique, et Sourisseau Jean-Christophe, ‘Les échanges sur le littoral de la Gaule méridionale au premier âge du Fer. Du concept d’hellénisation à celui de méditerranéisation’, Archéologie des rivages méditerranéens (50), 2010, pp. 237-246.
8 Bats Michel, ‘Les Phocéens, Marseille et la Gaule (VIIe-IIIe s. av. J.-C.)’, Pallas. Revue d’études antiques, (89), 2012, pp. 145-156.
9 Geel van Bas, ‘La crise climatique de 850 avant notre ère’, in Des climats et des hommes, ed. J.-F. Berger, Paris, La Découverte, 2012, pp. 257-268.
10 Favory François, Girardot Jean-Jacques, Nuninger Laure, Tourneux François-Pierre, ‘Archaeomedes : une étude de la dynamique de l’habitat rural en France méridionale, dans la longue durée (800 av. J.-C.-1600 ap. J.-C.)’, AGER, (9), 1999, pp. 15-35.
11 Nuninger Laure, ‘Peuplement et territoires protohistoriques du VIIIe au Ier s. av. J.-C. en Languedoc oriental (Gard-Hérault)’, Thèse de doctorat en Archéologie, Université de Franche-Comté, 487 p, 2002 [En ligne] https://halshs.archives-ouvertes.fr/tel-00002981/
12 Garcia D., La Celtique méditerranéenne 2nd ed., op.cit.
13 With more than 20 ha, the oppidum of Carsac at Carcassonne remains one-of-a-kind (Guilaine et al.1986).
14 Beylier Alexandre, Gailledrat Eric, ‘Traditions indigènes et innovations dans les fortifications de l’aire languedocienne à l’âge du Fer: l’exemple de Pech Maho (Sigean, Aude) et du Cayla de Mailhac (Aude)’, Revista d’arqueologia de Ponent, 19, 2009, pp. 251-270.
15 Gascó Jean, ‘La question actuelle des fortifications de la fin de l’âge du Bronze et du début de l’âge du Fer dans le midi de la France’, Documents d’archéologie méridionale. Protohistoire du Sud de la France, (32), 2009, pp. 17-32.
16 Nuninger L., Peuplement et territoires protohistoriques..., op.cit.
17 If one excepts the site of Saint-Blaise, for which the estimations are not very reliable from the fifth to the third centuries (Isoardi 2008 et 2013).
18 Nuninger Laure, « Pôles de peuplement et organisation territoriale au second âge du fer en Vaunage, Vidourlenque et Vistrenque (Gard) », Territoires Celtiques: espaces ethniques et territoires des agglomérations protohistoriques d’Europe Occidentale, Actes du colloque international de l’AFEAF, 2002, p. 139-149.
19 Garcia Dominique, ‘Agglomérations et territoires aux Ve-IVe s. av. J.-C. dans l’interfluve Aude-Hérault: proposition d’analyse’, in Cité et Territoire, ed. M. Clavel-Lévêque, R. Plana-Mallart, Besançon, ALUFC-Les Belles Lettres, 1995, pp. 175-186.
20 The archaeological indicators are very clear from this point of view, revealing a system of production firmly anchored in the territory with a developed agricultural practice implying varied production and special treatment for preserving the different products (Bouby 2003). Fishing zones and the processing of fish for preservation are also attested (Chausserie-Laprée 2005).
21 This idea remains very much subject to debate due to study of the archaeo-botanical remains, which tend to show year-round occupation of certain sites considered as seasonal by archaeologists. The rarity of the indicators and of the sites concerned does not at present permit acceptance of one or the other hypothesis (Boissinot 2014).
22 Nuninger L., Peuplement et territoires protohistoriques..., op.cit.
23 Arcelin P, ‘Entre Salyens, Cavares et Volques: peuplement protohistorique et dynamique culturelle dans la région Alpilles-Montagnette’, in Les Alpilles et la Montagnette, ed. F. Gateau, M. Gazenbeek, Paris, Académie des Inscriptions et Belles Lettres, 1999, pp. 61-78.
24 Lachenal Thibault, ‘L’âge du Bronze en Provence: bilan de cinquante ans de recherche’, ed. D. Garcia, Archéologie des rivages méditerranéens: 50 ans de recherche, 2010, pp. 223-235.
25 Nuninger L., Peuplement et territoires protohistoriques..., op.cit.
26 Feugère Michel, Py Michel, ‘Les monnaies préaugustéennes de Lattes et la circulation monétaire protohistorique en Gaule méridionale’, Revue numismatique, 6 (164), 2008, pp. 485-489.
27 Bouby Laurent, Garcia Dominique, ‘L’économie agricole à l’Âge du Bronze en France méridionale Apports récents de la carpologie’, in L’âge du bronze en Méditerranée Recherches récentes, Paris, Editions Errance, 2011, pp. 101-114.
28 In is, in fact, entirely probable that at this epoch property was markedly different from the land ownership model as known at least from Antiquity. Testart 2005, Chouquer 2010.
29 Buxo i Capdevila Ramon, ‘Evidence for vines and ancient cultivation from an urban area, Lattes (Hérault), southern France’, Antiquity, 70(268), 1996, pp. 393-407.
30 Py Michel, Buxo i Capdevila Ramon, ‘La viticulture en Gaule à l’Âge du Fer’, Gallia, 58 (1), 2001, pp. 29-43.
31 Nuninger L., Peuplement et territoires protohistoriques..., op.cit.
32 Trément Frédéric, Archéologie d’un paysage, les étangs de Saint-Blaise (Bouche du Rhône), Maison des Sciences de l’Homme, Paris, 1999
33 Chausserie-Laprée Jean, Martigues, terre gauloise: entre celtique et méditerranée, Éditions Errance, 2005
34 Py Michel, La sculpture gauloise méridionale, Paris, Éditions Errance, 2011.
35 Roure Réjane, Pernet Lionel, Des rites et des Hommes, Paris, Editions Errance, 240 p, 2011.
36 Verger Stéphane (ed.), Une Odyssée gauloise: parures de femmes à l’origine des premiers échanges entre la Grèce et la Gaule, Paris, Éditions Errance, 2013.
37 The numbering refers to the relations figured in the arrow diagrams of figure 3 – A for the first regime and B for the second.
38 The appearance of this is considerably prior to the period under consideration (Saulieu et Testard 2013).
39 In the sense meant by D. Garcia (2004): the man of highest status or having the most power within a group (leader) in a society which is not extensively hierarchised.
40 Par des sculptures de têtes coupées puis de guerriers (Py 2011).
41 Selon D. Garcia (2004).
42 Crabtree Stefani A, ‘Simulating Littoral Trade: Modeling the Trade of Wine in the Bronze to Iron Age Transition in Southern France’, Land, 5 (1), 2016, p. 5.
43 Except for Lattes, where the same process occurs later (Py 2009).
44 On whom it had relied, however, during its establishment (Sourisseau 1997).
45 Uri Wilensky 1999-2016, https://ccl.northwestern.edu/netlogo/
46 D’Ercole Maria Cécilia, Histoires Méditerranéennes, Paris, Éditions Errance, 223 p, 2012.
47 Gras Michel, Trafics tyrrhéniens archaïques, 258, Rome, Italie, École française de Rome, 1985.
48 The energy expressed in hl refers to the assistance provided by other individuals, certain historians having suggested the existence of festivals organised specifically to attract the aid of others at harvest time (Dietler 1990).
49 Py M., Culture, économie et société, op.cit.
50 Nuninger L., Peuplement et territoires protohistoriques..., op.cit.
51 The preference values are scaled from 0 ‘entirely in favour of the Etruscan wine’ to 100 ‘entirely in favour of the Greek wine from Marseille’, with 50 corresponding to an equal preference value for the Etruscan and Greek wines.
52 For a more thorough discussion of the results, see Crabtree 2016.
53 Ferber Jacques, Stratulat Tiberiu, Tranier John, ‘Towards an integral approach of organizations in multi-agent systems’, Handbook of Research on MultiAgent Systems: Semantics and Dynamics of Organizational Models, 2009, pp. 51-75.
54 Ferber Jacques, ‘Concepts et méthodologies multi-agents’, dans Modélisation et simulation multi-agents, applications pour les Sciences de l’Homme et de la Société, ed. F. Amblard, D. Phan, Paris, Hermès-Lavoisier, 2006, pp. 23-48.
55 Ferber Jacques, Dessalles Jean-Louis, Phan Denis, ‘Emergence in agent based computational social science: conceptual, formal and diagrammatic analysis’, Intelligent complex adaptative systems, 2008, pp. 255-299.
56 Wilber Ken, A Theory of Everything: An Integral Vision for Business, Politics, Science and Spirituality, Boston, Shambhala, 189 p, 2001.
57 Evolution of its characteristics over time.
58 The societies studied here are definitively agro-pastoral, even if, in this regime, they could have supplemented their food resources by gathering, hunting and fishing. In the model, we have chosen to employ the term ‘food gathering’ in a broad sense, including both the activities of gathering wild food resources (berries, roots, mushrooms…) and the activities of cereal-gathering as a practice for those agents who cannot ensure their own crops because of lack of space (fields) or reserves (seeds).
Auteurs
Chrono-Environnement UMR 6249, MSHE C.N. Ledoux USR 3124, CNRS/Univ. Bourgogne Franche-Comté
UMR 5140 Archéologie des sociétés méditerranéennes, CNRS/Université Paul Valéry Montpellier 3
UMR Espace Dev, IRD/ Université de Montpellier
(†) UMR 5140 Archéologie des sociétés méditerranéennes, CNRS/Université Paul Valéry Montpellier 3
UMR Espace, CNRS/Université d’Avignon
Le texte seul est utilisable sous licence Licence OpenEdition Books. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.
Quatre ans de recherche urbaine 2001-2004. Volume 2
Action concertée incitative Ville. Ministère de la Recherche
Émilie Bajolet, Marie-Flore Mattéi et Jean-Marc Rennes (dir.)
2006
Quatre ans de recherche urbaine 2001-2004. Volume I
Action concertée incitative Ville. Ministère de la Recherche
Émilie Bajolet, Marie-Flore Mattéi et Jean-Marc Rennes (dir.)
2006