Contribution of the Microhistorical Approach to Landscape and Settlement Archaeology: Some French Examples

2.1.1 For an Approach to the “Everyday” Spatial Practices of Ancient Societies

The object of study being the relations between successive societies and their space, the anthropization of the environment and its appropriation by Man, it is necessary to place oneself at the scale of almost daily intervention of past societies. This scale is that of close movements: from one village to another, from the village to the hamlet, from the hamlet to the field, from the field to the forest. The object of study is the finage, a form of appropriation of space through the definition of agrarian territories, probably as early as Protohistory. It is a question of restoring the rhythms of appropriation of space by ordinary rural communities.

A fine spatial resolution also makes it possible to take into account an extended chronological scale, and this is necessary insofar as the imprint of Man on the environment is legible at least from the Neolithic period. The challenge is to take advantage of a small study area where information – particularly archaeological information – is well controlled in order to measure the rates of evolution of the various components of the anthropized space: settlement, road network, land parcels, and resources. This is from Protohistory to the nineteenth century.

2.1.2 For Optimal Data Collection Resolution

Taking into account the long term implies the integration of information delivered by different sources. The choice of the micro-scale provides the ideal framework for their cross-referencing and comparison and makes it possible to enrich the results obtained from each of them. A micro-regional study mobilizing sources of various kinds favours different perspectives on the same object of study: the formation and transformations of the landscape. The contribution of an archaeological field survey is decisive because it is often the only source documenting the periods prior to the eleventh century. Our knowledge of the settlement and occupation of the land from Protohistory to the early Middle Ages is directly linked to the quality and scope of the ground survey operations. The micro-scale makes it possible to subject the archaeological data collected to very detailed questioning, linked to agrarian practices for example, where they can be compared with other sources (archives or physicochemical data).

If archaeological data are rare and acquired at the cost of a lot of time, written sources are directly available and in increasing quantity as time goes by. The micro-scale therefore allows for a thorough mobilization of these data: systematic recording of all mentions of topographical elements, approach to agrarian territories, restitution of the limits of immaterial territories (courts, tithes, parishes) that also contribute to the construction of space, which would be impossible on a smaller scale. Finally, the micro-scale is that of the survey and the initial use of the old plans integrated. The old cadastre is constructed by the commune, composed of sheets surveyed at 1:2,500 or 1:5,000. The resolution of the information allows us to study the basic unit of territorial division: the parcel. If we consider the parcel of land as a possible witness of the land development of the past, it is at the scale of the cadastral survey that we must place ourselves. The same is true for the older plans that are also taken into account: the “terrier plans” surveyed on the scale of a landlordship can be used at scales of between 1:2,000 and 1:5,000. They can be directly compared with the old cadastre.

At the micro-regional level, the ambition to analyse all the components of the space appropriated by ancient societies therefore requires all available sources to be taken into account. This means drawing on a multitude of disciplines, from geography to history, each with its own historiographical traditions and methodological specificities. The modalities of this interdisciplinary dialogue are a challenge in themselves (Chouquer & Watteaux, 2013).

2.2.1 Collective Research Projects on the Rural Settlement of the Early Middle Ages

If archaeological excavation, for a long time confined to the strict confines of residential buildings, forced the medieval archaeologist to carry out a forced micro-analysis, the development of preventive archaeology in the 1980s and especially the 1990s allowed the accumulation of numerous field data concerning rural settlements of the early Middle Ages. The first few sites, excavated between the 1960s and 1980s, contributed to illustrating the miserable image of the Alto-medieval period based on the scarcity of written sources. The structures made of perishable materials and the short duration of occupation allowed the reconstruction of precarious and ephemeral settlements, which were supposed to testify to the poverty of agrarian techniques and the weakness of the demography, in a supposed context of generalized crisis. In the 1990s and 2000s, the series of these numerous habitats within the framework of collective research projects made it possible to read the typology and chronology of land occupation in the Early Middle Ages on a regional scale. The publication of synthesis of these different projects then made it possible to appreciate the common trends and local particularities in the methods of implantation, structuring, and dynamics of these settlement networks (Faure-Boucharlat, 2001; Gentili, Lefèvre, & Mahé, 2003; Le Boulanger, 2023; Lorren & Périn, 1995; Peytremann, 2003; Valais, 2012).

In this case, the microhistorical analysis linked to the excavation and interpretation of each individual site could be overcome in order to build models of land use and settlement dynamics valid at higher scales (Section 2.3).

2.2.2 Archaeomedes and Archaedyn

At the micro-regional scale, as we have seen above, the spatial micro-scale is first of all a methodological choice allowing a balance between the different categories of sources mobilized to reconstruct landscape dynamics: archaeological sources acquired at the cost of considerable fieldwork, written and planimetric sources that are abundant as the focus is widened (Poirier, 2021).

Here again, if the micro-historical approach has made it possible since the 1980s to read, in the long term permitted by the spatial micro-scale, the first spatio-temporal trajectories of rural spaces developed by ancient societies (Astill & Davies, 1997; Buchsenschutz, Coulon, & Gratier, 1988; Poirot, 1998; Zadora-Rio, 1987), the comparative approach has long been made difficult both by the rarity of this type of approach and by the methodological biases linked to the acquisition and processing of data: variable field procedures in pedestrian survey, quality and quantity of written sources, varied political and natural contexts. These obstacles have been partially overcome and quantitative and comparative approaches have been developed in the framework of two collective research projects carried out between 1990 and 2010: Archaeomedes and Archaedyn.

The European Archaeomedes program marks a step forward in the approach to the relationship between landscape and settlement by taking into account a broad chronological spectrum running from 800 BC to 1600 AD and a systematic study of the relationships maintained between habitats and their environment and of the respective evolutions of each of the components of the system. Its objective is to study “the interactions between social dynamics on the one hand and natural dynamics on the other”, in the long term, in order “to elucidate the complex dynamics of land degradation […] in vulnerable circum-Mediterranean environments. It focuses in particular on the relationships between the natural processes involved and the socio-economic dynamics underlying human interaction with the environment” (Van Der Leeuw, Favory, & Fiches, 2003, p. 11). The French team involved in this program mobilized the data available in the lower and middle Rhone Valley and compared them with environmental data. This has made it possible to demonstrate that the changes that occurred in the density of the habitat in the second–third century AD cannot be interpreted as a depopulation of the countryside and that the reorganization of the habitat networks bears witness more to a process of “rationalization of the production system” than to a change in climate or soil exhaustion (Van Der Leeuw et al., 2003, p. 202). This program is not only one of the first experiments in France to model archaeological data on a regional scale for historical periods, but also one of the first attempts to model settlement dynamics using archaeological data, written sources, and maps. The Archaeomedes program marks a step forward, in particular by the definition and implementation of new methodological tools and descriptive criteria of rural settlement dynamics. Examples include the calculation of the ratio of the number of created sites to the number of abandoned sites to measure the weight of each time step, the degree of anthropization of the surrounding environment (measured by adding the occupation periods of all the settlements occupied prior to the creation of the site within a radius of 500 m), or the weighting for each period of the number of sites by the total area occupied. Many of these indicators are now routinely used in rural settlement studies.

The ArchaeDyn programme, “Spatial Dynamics of Territories from Prehistory to the Middle Ages” was conducted in two phases: between 2005 and 2007 (Gandini, Favory, & Nuninger, 2012) and between 2009 and 2012 (Gauthier, Georges-Leroy, Poirier, & Weller, 2022). It contributed to broadening the research problem initiated by the European programs Archaeomedes I and II (Durand-Dastès et al., 1998; Van Der Leeuw et al., 2003), by approaching the dynamics of territories, considered from several new angles of study, both thematically, chronologically and spatially. It was conducted by a group of nine teams from six “Maisons des Sciences de l’Homme” (Besançon, Clermont-Ferrand, Dijon, Nanterre, Nice, Tours). The aim of the project was to deepen the analysis of spatial interactions between social groups, the resources they exploit to ensure their subsistence, to produce objects and to exchange with other groups at short, medium, and long distances, and the itineraries that govern the flows at different periods. The collective work was structured along three thematic axes – 1. “Supply areas, terroirs and finages”; 2. “Settlement, networks and territories”; 3. “Circulation of raw materials and objects” – complemented by an axis 4 dedicated to “Methods and tools for spatial analysis” (Saligny et al., 2012). Compared to the analyses centred on habitat sites developed in the Archaeomedes I and II programs, the Archaedyn collective has instead focused on the factors that contribute to producing a more or less dynamic environment according to land use modalities. Thus, for all of the thematic axes, the aim was to answer three major questions: what spaces are permanently occupied? Which spaces were conquered and then abandoned? Which spaces were used regularly but never permanently? The comparative, spatial, and diachronic approach required the design and development of indicators whose calculation had to be strictly reproducible across study areas, despite the apparent heterogeneity of the field survey methods and the scientific, historical, and environmental contexts specific to each. The study of agrarian manurings puts the weight of socio-environmental constraints into perspective, showing the capacity of ancient societies to overcome them in order to expand their agrarian space despite, for example, the agro-ecological limits of soils, topographical conditions, or the distance from settlement centres. Variations in the extent of amended spaces and the intensity of amendments over the long term define the trajectory of rural space development, which is complementary to that of the settlement system. This examination leads to revisit certain historiographical topoi such as the massive and generalized impact of Roman colonization in terms of rural development, the “abandonment” of Late Antiquity, or the economic sluggishness of the High Middle Ages (Section 2.3). The comparison of study areas, both northern and southern France, illustrates the diversity of situations and compensates for the very local vision of the analysis of the terroir.

2.3.1 The Transition From Antiquity to the Middle Ages

Initially shaped by the scarcity and paucity of written sources, the idea of a break between Antiquity and the Middle Ages was also based on the observation of the change in construction methods, from the mortar and stone of the Roman period to the perishable materials of the architecture of the High Middle Ages. This architecture of perishable materials that is difficult to read was initially interpreted as an indication of the ephemeral nature of the occupation. But the vestiges observed do not allow us to judge the quality of the constructions, which can be very well done, even in earth and wood. The supposedly ephemeral nature of the settlements has influenced the vision of the agrarian economy of the early Middle Ages, particularly in relation to the continuity of present-day villages. This difference was attributed to a supposed failure of early medieval agriculture, where the exhaustion of the soil would have led to the abandonment of settlements to cultivate new land. In reality, we will see that this ephemeral image was linked to the small size of the excavated surfaces, which did not allow us to see a form of continuity of occupation, but with small displacements.

In Brittany, there is often a continuity of occupation between Antiquity and the Middle Ages (Catteddu & Le Gall, 2023; Le Boulanger, 2023). Small farms and villas remained in place, even though facilities such as thermal baths, which were costly to maintain, were quickly abandoned or converted. In some cases, wooden buildings, often small-scale, were built around the earlier domains. Trade routes were maintained, particularly with the British Isles and northern Europe. From the sixth century onwards, there appears to have been a break with the past. The earlier settlements were gradually abandoned, and new areas were created in cultivated zones. The settlements uncovered in the east of the peninsula correspond to farming units set in regular quadrangular plots, often using a plot grid inherited from Antiquity. In the west of the peninsula, settlements were built in circular, oval, or horseshoe-shaped enclosures. They were either scattered or grouped together in small hamlets, often reusing the ancient plots and roads.

In the north-west of France (Valais, 2012), the analysis of the chronological distribution of settlements excavated in rescue archaeology shows a low representation of the occupations of the late antiquity, a significant increase in the number of sites around 600 and especially from 700 to relapse at the end of the millennium, and the existence of late sites implanted ex-nihilo in the eighth or ninth century and lasting until the twelfth or thirteenth century, the fruit of chance or places isolated from the ancient territories and implanted during phases of Carolingian growth or the Central Middle Ages.

In central France (Nissen, 2014), the transition between Antiquity and the Early Middle Ages remains difficult to understand, particularly because of the difficulty in identifying fourth-century occupations. The occupations of the fifth-sixth centuries, on the other hand, are well represented in both the preventive excavations and the territorial surveys. The identification of these phases of occupation has been made possible by the progress made in the study and dating of ceramic production in this region (Husi, 2022). At the scale of the settlements, most of the occupations lasted two or three centuries, but the excavations that revealed a duration of four centuries or more are almost as numerous. This apparent stability does not, however, reflect a topographical anchorage: on closer inspection, the longest occupations reveal slight shifts in the habitat. The sites are constantly evolving, their spatial organization changes over time, and certain phases are often distinguished by a particularly dense occupation. This raises questions about the precise meaning of dating, based on ceramics with chronological margins that are often of the order of one to two centuries, and about the scope of the excavations, which rarely cover the entirety of the habitats. This mobility, characteristic of early medieval sites, has long been wrongly considered a sign of weakness in ancient research. In reality, this mobility testifies to a constant interaction between the habitat and its environment.

In the Lyon region (Faure-Boucharlat, 2001), gallo-roman material is frequently discovered on early medieval habitat sites, but it is qualified as residual or erratic by the excavators, because it cannot be associated with built remains. It is difficult to distinguish between a real continuity of occupation of the settlements after the Late Antiquity Period and the resumption of occupation after a phase of abandonment. Sometimes we observe a form of determinism in the spatial organization of the settlements, which suggests a relationship with the ancient plot of land. The most frequent situation is the installation of communities in ignorance of what came before. The earliest of these settlements created ex nihilo begin at the end of the fifth century, but for the most part, the occupation is centred on the seventh century. The Carolingian period is particularly poorly represented, whether one considers the settlements that appeared at the beginning of the Middle Ages or those that took over from older settlements, with or without an interruption in occupation. The turn of the millennium is manifested by a densification of the occupation. Some settlements were established in places that had been unoccupied during the early Middle Ages and were considered to be new settlements. The phenomenon is not only reflected in creations ex nihilo but also affects sites that had been occupied in the early Middle Ages. It is indeed in the middle of the Middle Ages that the densest and most structured phase of occupation occurs.

In the south-east region (Schneider, Fauduet, & Odenhardt-Donvez, 2007), on the threshold of the fourth century, the history of the dynamics of settlement and the transformations of rural habitat in Mediterranean Gaul is first placed in a long movement of decrease of the number of occupied sites. The phenomenon is not sudden but stems from an earlier shock wave, the first signs of which are perceptible from the end of the first century, and become considerably more pronounced in the second half of the following century and the third century. Two systems of settlement would be opposed to each other. The centrifugal system of the abundance of creations during the brutal growth phase of the High Empire, which invested the space around poles – agglomerations or domains – and the centripetal system of the Late Empire, in which the rural economy was refocused around less numerous but more grouped settlements. The movement of creation of new sites, which begins to be perceived from the fourth century onwards, becomes more and more evident in the course of the fifth century, especially at the junction of the fifth and sixth centuries. On the threshold of the sixth century, the situation appears to be clearly different from what it was a century earlier. The emergence of new rural centres on the margins and the restructuring that can be seen in areas of ancient settlement, such as Lunel-Viel (Garnier, Legrand, Mercier, & Raynaud, 1995; Raynaud, 2007), are the first signs. Other signs are the significant increase in the number of new sites and a wider occupation of the space. The analysis of the chronological distribution of the surveyed and excavated sites highlights the essential role of the sixth century as a threshold of change in the domanial landscape, while suggesting that these transformations are far from having completely erased the topographical stability of the network of domanial centres in the first half of the first millennium.

The vision of a High Middle Ages in the continuity of Antiquity was forged quite early by the written sources (with the integration of many Germans among the Roman elites), but the archaeological data indicate a break that can be seen through the abandonment of many villae in the third century. However, a greater continuity of ancient structures can be observed in the south of France. Finally, regional differences lead us to subdivide the Early Middle Ages into two periods: fifth–seventh centuries and then seventh–eleventh centuries. The hinge of the seventh century corresponds to the conjunction of several factors that announce the beginning of a growth cycle: population growth, diffusion of new techniques, organization of work, trade, and urban life. These chronological revisions, made possible by a series of micro-scalar archaeological observations, have recently led to the proposal of a new chronological division of the Middle Ages which places less importance on a supposed break in the tenth-eleventh centuries and more on a better characterization of these processes over the long term (Mazel, 2021).

2.3.2 Polarization: Towards the Village

Just as Gallo-Roman archaeology has long concentrated on the study of the residential parts of the villae, so the archaeology of the medieval rural habitat, dominated by the problems defined by the written sources, was mainly interested in the phenomenon of deserted villages at the end of the Middle Ages (Beresford & Hurst, 1990; Demians D’Archimbaud, 1980). Then, following the evidence of the great mobility of the medieval rural habitat by surveys and especially preventive archaeology work, the interest of medieval historians and archaeologists shifted from the late Middle Ages to the periods of the tenth–twelfth centuries, following the thesis of P. Toubert describing the incastellamento model (Toubert, 1973). Developed from written sources, it expresses the transformation of the habitat and the reorganization of the land induced by the establishment of feudal society. It emphasizes the role of the castle in the concentration of the habitat. Robert Fossier, while expanding the model (from the incastellamento to the encellulement), notes that in most regions of Europe, between the tenth and twelfth centuries, territorial structures – lordships and parishes – were set up to provide a framework for society on a micro-scale (Fossier, 1982). Thus, the numerous abandonments of dispersed settlement sites around the year 1,000 would be part of a process of nucleation in favour of the ancient villages that we still know today but which are very rarely the subject of archaeological surveys.

From the 1990s, French medieval archaeologists began to question the possibility of archaeologically documenting the “birth” of the village, given the diversity of scenarios that were becoming apparent as a result of the increasing number of excavations of early medieval rural settlements. First of all, they pointed out the need to distinguish between the “physical” village accessible to archaeology and the “social” village described by written sources (Zadora-Rio, 1995). Second, a close examination of the chronologies of the creation and abandonment of rural sites made possible by preventive archaeology has enabled us to nuance the role of the turn of the millennium in the evolution of the settlement pattern (Peytremann, 2003, 2013). The most recent positions rightly call for the historical object of the “village” to be reintroduced into a broader study of the forms and dynamics of rural settlement between the fifth and fifteenth centuries (Catteddu, 2021; Catteddu, Raynaud, Le Goff, & Chauveau, 2015; Lauwers, Watteaux, & Catteddu, 2021; Peytremann, 2014; Watteaux, 2003).

If the search for a “birth of the village” is illusory, the fact remains that the polarizing role of the church was manifested in the first stage by the progressive grouping of burials within the framework of the affirmation of the parish and its Christian cemetery. This has been particularly demonstrated on a micro-scale by the monographic approach of ecclesiastical and cemetery sites such as the emblematic cases of Rigny-Ussé in Touraine or Vilarnau in Catalonia.

At Rigny-Ussé (Indre-et-Loire, Center of France), the excavation of the area around the church conducted between 1986 and 1999 revealed a very long period of occupation of the site, from the seventh century in the excavation area – and from the High Empire in the immediate environment – until the middle of the nineteenth century. During the first period of occupation, in the seventh and eighth centuries, despite the existence of a place of worship at Rigny, the inhabitants of the colony were buried elsewhere, probably in a necropolis located at a distance. From the middle of the eighth century onwards, burials gradually occupied the ruins of the buildings and the spaces in between. The burial area is very sparse and extends along a north–south axis. The church, near which burials were made, does not seem to be a focal point for burials. The construction of a new church on the same site at the end of the tenth or beginning of the eleventh century coincides with a first reduction in the burial area, which corresponds to a reorganization of the burial space around the church. This development is not the result of external pressure, since the space freed by the shrinking of the burial area was not occupied by buildings until two centuries later. Nor is it due to a decrease in the burial population, since the reduction of the burial area went hand in hand with an increase in the density of burials. It should therefore be seen as a stage in the establishment of the parish organization marked by a new interdependence of church and cemetery. At the beginning of the twelfth century, a new reduction of the burial space is accompanied by the construction of a fence wall which materializes the limit of the cemetery (Zadora-Rio et al., 2020).

The church of Vilarnau (Pyrénées Orientales, Catalogne) was probably already built at the end of the ninth century and the dimensions given to it at that time did not change until its destruction at the end of the eighteenth century. At that time, the church seemed isolated in the middle of the plateau, but it quickly polarized a first cemetery that extended around the building, the furthest tombs never being more than 30 m away. The immediate environment of the church of Vilarnau evolves radically during the second half of the tenth century or at the very beginning of the eleventh century with the temporary installation of a habitat on the bangs of the cemetery. This grouping around the church coincides with the abandonment of many nuclei of settlement – isolated farms or hamlets – located in the nearby countryside. The habitat that settled around the church faded away before the twelfth century. The cemetery and the church continue to be used but are not sufficient to fix the habitat in their immediate contact to the benefit of regrouping around the castle, located only 500 m to the northeast. The fact that the building continues to be maintained and that its cemetery is used shows that there is no break, but a redistribution of the habitat within a territory that probably already forms a parish (Passarius, Donat, & Catafau, 2010; Passarrius, Donat, & Catafau, 2008).

2.3.3 Medieval Agrarian Growth and Decline

The scarcity of excavations of rural settlements from the Early Middle Ages until the 1980s led to the interpretation of this period as a phase of the crisis in the countryside, characterized by precarious and ephemeral settlements linked to failing agriculture. However, historians of written sources such as G. Duby had already identified signs of growth from the eighth century onwards through an apparent retreat of the forest, a multiplication of settlement sites, or a quantitative reading of written sources, all of which indicated a certain demographic growth during the last centuries of the first millennium.

Within the framework of the Archaedyn program described above, we have for the first time compared the trajectories of agrarian development of areas located in both northern and southern France. For this purpose, we have developed a quantitative and comparative approach to the remains of manuring revealed by the material collected off-site during systematic pedestrian surveys. The calculation, for each workshop area and for each period, of the sum of the surveyed plots having delivered at least one evidence of amendment (at least one shard), makes it possible to read the pulsations of the amended spaces in the long term in terms of extension or retraction. If we observe everywhere an extension of the amended spaces during the Roman period, we also record everywhere a certain abandonment during the late Antiquity. In three cases, the decrease in amended areas continued during the High Middle Ages. The developments are very diverse for the Central Middle Ages. Although three workshop zones show approximately the same score around one-third of the amended area, this proximity masks real disparities in the trends observed. Among the two workshop areas that had shown a significant increase in the amount of land amended during the previous phase, one saw this confirmed by a new increase in the amount of land amended, while the other showed a new decrease in this area. On the other hand, the three workshop areas that had been abandoned in the early Middle Ages show a more or less strong recovery in the central and final Middle Ages.

This study thus makes it possible to document the debated question of the existence and modalities of a certain form of agricultural growth in the early Middle Ages (Devroey, 2008; Leturcq & Mazel, 2021). The various readings and re-readings of Carolingian cartularies and polyptychs crystallized the debate in the 1980s (Centre culturel de l’abbaye de Flaran, 1990). While data from preventive archaeology now provide occasional information on this question – by uncovering areas for the silage of agricultural products (Carme & Henri, 2010), organized plots of land, and numerous agricultural establishments (Carré, 2011; Faure-Boucharlat, 2001; Gentili et al., 2003; Peytremann, 2003) – the estimated surface area occupied by cultivated areas is a key piece of information for assessing this growth. It is very clear in Touraine and Berry, where the increase in amended areas is spectacular; it is much more difficult to identify in southern contexts. This study has finally demonstrated the capacity of ancient societies to overcome certain geographical constraints. If the growth of cultivated areas is made up of successive takeovers and abandonments, it is evident over the long term. The production needs of agricultural commodities have led successive societies to integrate into their cultivated areas geographical contexts that are a priori unfavourable: steep slopes or those with little sunlight, heavy or acidic soils, and marginal areas far from inhabited areas. This integration had to be accompanied by technological progress and additional efforts in terms of soil improvement, as shown by the increase in the density of manure-spreading material. The comparative approach makes it possible, not to question existing macroeconomic models in their entirety, but to observe their local modulations by going down to the micro-regional level and varying the contexts of observation.