Éléments de géographie humaine et économique à l’âge du Bronze et au premier âge du Fer dans la plaine de Troyes

EnglishThe recent intensification of construction work and resulting preventive archaeological excavations on the Troyes Plain (Aube, France) have produced a remarkable increase in the amount of data available, considerably modifying our vision of early human occupation there. For some previously poorly known archaeological periods the density of sites with abundant finds raises a number of questions that surpass the individual interest of each site. This major renewal of the available evidence particularly concerns the late Bronze Age and the Early Iron Age. Most data relate to settlements and their development, thus providing an important body of evidence for gaining a broader understanding, albeit indirect, of the whole pattern of settlement. Analysis of the settlement system, coupled with the main data on the agricultural economy, was undertaken on two scales: on the large scale of a group of terroirs, using results of preventive excavations in the Parc Logistique de l’Aube (PLA), and on the smaller scale of the Troyes Plain, over a zone of about 15 km radius. This dialogue enables the regularities and irregularities in the overall settlement system to be detected and discussed, and local scenarios of change to be proposed. The dataset from the Parc Logistique de l’Aube is sound and representative. The spatial extent of the project (250 hectares) and the numerous discoveries dated to these periods allow a quantified vision of both the tempo of site occupation and changes in the agrarian system. Study of the settlement system was based on 25.5 hectares of extensively excavated zones investigated between 2005 and 2008. These include 84 spatially and chronologically distinct occupation sites, defined on the basis of 437 reliably dated archaeological features. Three generic indicators of their main dimensions (surface area, number of features, storage feature volume) were recorded for these occupations and are presented in histogram form with a chronological resolution of 50 years. Certain categories of data such as those relating to the plant and animal economy, for which each phase produced a significant number of samples, can be considered representative not only at the terroir level but also on the scale of southern Champagne. As regards the settlement system, the short-lived episode of a clustering of population (hamlet or village) at the end of the Bronze Age contrasts with the standard pattern of scattered protohistoric dwellings. The analysis shows six major phases in the development of the system and these are also visible throughout the Troyes basin: from − 2200 to − 1350, from − 1350 to − 950, from − 950 to − 750, from − 750 to − 620, from − 620 to − 450 and lastly from − 450 to − 250. On the smaller scale, several poles of high density stand out; they could indicate the emergence of a territorial structure that subsequently disappeared, during the later Iron Age. In addition, the agropastoral economy and its development provide evidence for the economic life and the capacities for reproduction of past societies. The dynamic nature of agriculture is attested by a large quantity of evidence, including a wide range of plants. The system is based on 15 cultivated species dominated by the production of cereals, with relatively few changes over time. While the range of crops and the economic importance of each plant show little variation, a change can be seen in the wild plants associated with the remains of the domestic species. This is an indication of changes in agricultural practices. The faunal evidence follows a rather different pattern, as data are abundant for the end of the Bronze Age and this is followed by a period between − 750 and − 550 with very few faunal remains. For the end of the Bronze Age, there are no special trends in consumption, herds are made up of the usual domestic triad and there are very few wild animal bones. The final discussion addresses a number of open questions. What is the nature and value of the economic cycle identified through these quantitative variables? Can one go beyond the hypothesis of the fundamentally scattered nature of the protohistoric settlement system, as one observes a slow but irrepressible trend towards more long-term occupation through a reduction in mobility between generations? On a similar note, the presence of short-lived villages at the end of the Bronze Age provides additional evidence for considering the issue of possible social resistance to the constraints of urbanisation in continental Europe. Lastly, all these data also fuel the debate on the degree of resilience of protohistoric farming societies to the climate changes recorded for the end of the Bronze Age and the Early Iron Age. francaisL’intensification recente des travaux d’amenagement donnant lieu a des operations d’archeologie preventive dans la plaine de Troyes (Aube) a entraine une expansion remarquable du volume des donnees, contribuant a modifier nettement la vision que l’on pouvait avoir de l’occupation humaine ancienne. La densite et la richesse des sites de certaines periodes archeologiques jusqu’a present inegalement documentees suscitent de nombreuses questions depassant l’interet propre de chaque site. Ce renouvellement en profondeur de l’information touche particulierement la fin de l’âge du Bronze et le premier âge du Fer. Le cœur des donnees a trait au systeme d’habitat et a son evolution qui peut etre considere comme le principal indicateur utile a la comprehension, bien qu’indirecte, de la trame de peuplement ancienne. Le travail d’analyse du systeme d’habitat, couple aux donnees principales du systeme economique agropastoral, a ete mene a deux echelles : d’une part a grande echelle, celle d’un groupe de terroirs, par le biais de la documentation des fouilles preventives menees sur le parc logistique de l’Aube (PLA), d’autre part, a plus petite echelle, celle de la plaine de Troyes, sur une zone de 15 km de rayon environ. Ce dialogue permet de detecter et de discuter des regularites ou des irregularites dans la trame globale du systeme d’habitat et de proposer des scenarii locaux d’evolution. L’assise du parc logistique de l’Aube, en termes de representativite, fournit un socle de donnees solide. L’etendue spatiale de l’operation et les decouvertes nombreuses se rapportant a ces periodes autorisent une vision quantifiee et plus seulement qualitative des rythmes d’occupation de l’habitat comme de l’evolution du systeme agraire. Certaines categories de donnees telles que celles relatives a l’economie vegetale ou animale, dont chaque phase a livre un nombre consequent d’echantillons, peuvent etre considerees comme representatives non seulement au niveau du terroir etudie mais egalement a l’echelle de la Champagne meridionale. Du point de vue du systeme d’habitat, la courte experience d’un regroupement de population (hameau ou village) a la fin de l’âge du Bronze tranche avec la trame courante de l’habitat disperse protohistorique. L’analyse conduit a distinguer six grandes phases dans l’evolution du systeme, que l’on retrouve a l’echelle du bassin de Troyes. Se degagent egalement, a petite echelle, plusieurs poles de densite forte qui pourraient signaler une structuration du territoire qui a disparu ulterieurement, pendant le second âge du Fer. L’economie agropastorale et son evolution sont aussi sollicitees en tant que principal indicateur de la vie economique et des capacites de reproduction des societes passees. La discussion finale aborde quelques themes qu’il importera de preciser : sur la nature et la valeur du cycle economique identifie, sur la these du caractere fondamentalement disperse du systeme d’habitat protohistorique et de sa resistance a l’urbanisation, ainsi que sur la resilience des societes agropastorales face aux changements climatiques globaux enregistres a la fin de l’âge du Bronze et au premier âge du Fer.