Chapter 10: Transition 7: From the ancient to the medieval world (4th-8th centuries)
Entrées d’index
Keywords : agglomeration, Late Antiquity, context, ecclesiastic, hierarchy, modelling, power, regime, transition, Southern France
Texte intégral
Introduction
1The object of this transition is nothing less than the passage from one ‘historical regime’, that of the Mediterranean civilisation of Antiquity, to another regime which the Western historiographic, philosophical and religious tradition radically opposes to it, the regime of the Christian civilisation of the Middle Ages. By ‘historical regime’ we mean a construction founded on ample conceptual, documentary, even political, material, enshrined by a clear-cut division applied by academic disciplines.
2As much as, if not more than, other passages, this one raises the question of transition. The notion of transition, as it has been brought to bear by the TransMonDyn programme, is defined as the process leading from regime to another: how to define and grasp this if one considers that in reality, in history, everything is always in transition, in transformation? The notion of transition is obviously the corollary of another historical question, that of regime, considered as a stable state – something which presupposes an agreement to consider a state of things as fixed or in balance, having no relation to the entropy inherent within every socio-spatial system.
3After much reflection, we have chosen to treat the transition from late Antiquity as beginning in the fourth century C.E., when the imperial political regime of the Tetrarchs, then of Constantine, marks a significant rupture with the model of the Principate inherited from the reign of Augustus, and as extending to, and including, the eighth century C.E., when the Carolingian dynasty displaced that of the Merovingians (accession of Pepin le Bref in 751). The space of reference on which this transition is explored is the territory of the province of Narbonnensis prima, a province initially imperial, then Visigothic (figure 1), corresponding roughly to the Languedoc-Roussillon region, whose administration was the object of numerous power-struggles: power-struggles between the civitates1 and the Roman empire, then between the cities and the Germanic, Visigothic and Frankish kings, on the one hand, and the bishoprics, on the other hand: in this respect this space is emblematic of the transition under consideration.
4Moreover, the history of this part of the Narbonnensis region has profited from the work of Languedoc researchers such as, among others, Jean-Luc Fiches, Pierre Garmy, Christophe Pellecuer, Claude Raynaud and Laurent Schneider, who have made it possible to construct a structured and relatively well documented narrative of this transition. Such apparent coherence should not lead one to forget the contradictory and fragmentary documentary heritage, which prevents responses to a number of fundamental questions: in these conditions, it is advisable to restrict the size of the space under study by concentrating on a region and a history which are relatively well known.
5Transition 7 was initially conceived and approached less as an anticipation of Transition 8 (800-1100) than as a succession to Transition 6, which issued in the end of the settlement cycle corresponding to the Romanisation of the Gallic space and put in place the Gallo-Roman settlement system.2 The reason for this perspective was the historical and archaeological documentation available, which imposes constraints when it is fragmentary and discontinuous: an approach to the dynamic of the territories and the network of agglomerations was therefore privileged, in order to try to untangle the interplay of factors and territorial actors at the different levels of the entities implicated in the organisation and management of the space under study. In the second and third centuries, the settlement system undergoes a major restructuring and records a drastic reduction in the number of establishments, especially in the dispersed habitat3, but, to a lesser degree, this movement also affects the secondary agglomerations4. This constitutes regime 1, which opens the transition from late Antiquity, extending from the fourth to the sixth century, during which a profound reorganisation of the imperial system takes place, disrupted by the invasions of Germanic peoples, which create kingdoms, and by the durable implantation of Christianity, which inaugurates its own administrative geography. In parallel, the settlement system is modified. Regime 2, in the seventh and eighth centuries, sets the seal on the fragmentation of the ancient provincial space, with a dissemination of civil and religious seats of power and a relative mobility of the sites of power.
6We propose in this chapter to give an account of the process of passing from the structured narrative to the formal representation of the transition, which precedes the definition of an operational model. This passage necessitates intermediate steps. We present here its initial and final states: the initial state corresponds to a structured narrative, identifying and contextualising the salient facts; the final state is the initialisation of the formal expression of that narrative.
The transition from late Antiquity in Languedoc-Roussillon
7The Tetrarchy, at the end of the third century, heralds a significant rupture with the imperial model created by Augustus and his successors. The political, military, administrative and fiscal reforms of Diocletian, followed by those of Constantine, thoroughly modified the conception, organisation and management of the imperial space, whose administrative geography was re-arranged at all levels, up to and including that of the province.
8This transition deals with the evolution of the Western imperial space5 under the impact of the Tetrarchy’s reforms, then under that of the Germanic kingdoms in the fifth century. During this period, the imperial power encouraged and stimulated the diffusion and institutionalisation of Christianity, as evidenced by the creation of the bishoprics, beginning in the fourth century, whose territorial authority had to compromise with the civic territorial framework6 before entering into competition it.
9The issues involved in Transition 7, merely in terms of the organisation of the civic space, are multiple:
How did the re-arrangements of the Western imperial space take concrete form in the geography of provincial and civic power?
How did the provincial territorial structures evolve in the space of the Germanic kingdoms?
How did the structure of cities evolve in the context of the Germanic, Visigothic, then Frankish kingdoms?
How did the emergence of castra function in the structure of the civitates ?
How did the civic framework receive and adapt to the new territorial entity constituted by the Christian diocese?
How did the existing Gallo-Roman aristocracy accommodate itself to the Germanic aristocracy that established itself beginning in the fifth century?
The spatio-temporal framework of the transition
10The spatial framework chosen for modelling Transition 7 is that of Languedoc-Rousillon, enrolled within the province of Narbonnensis, reworked and subdivided by the terms of the reform of Diocletian. The cities of the region were then organised into the province of Narbonnensis prima, also referred to as Septimania7 or Gallia when it was integrated into the Frankish empire. At the beginning of Transition 7, which corresponds to late Antiquity8, the province of Narbonnensis prima comprised five cities: Narbonne, Béziers, Toulouse, Nîmes et Lodève, that is, two Roman and three Latin colonies9 (figure 1). At the end of the fourth century and beginning of the fifth, a castrum at Uzès is also recorded, either a new city or a potential city created to the detriment of Nîmes10.
Figure 1: The framework of cities in Gallia Narbonnensis and its evolution in the sixth century, with the creation of bishoprics.

(Source : Schneider 2008a).
11Moreover, the new province had lost two Latin colonies, Ruscino (Château-Roussillon), in Roussillon, and Carcassonne, in western Languedoc. The latter is mentioned as a castellum in the Itinerarium Burdigalense (Itinerary from Bordeaux to Jerusalem, 333). As for Ruscino, already in decline in the first century, it ceded its place to the castrum of Elena (Elne, ancient Illiberis), attested in 350. These two down-gradings reinforced the city of Narbonne.
12During virtually the whole period of the transition, this province, centred on the ancient provincial capital of Narbonne, belonged to the Visigothic kingdom – from 418 to 711, at which date the Islamic conquest took possession of this essentially Spanish kingdom, its last refuge in Gaul. In 759, the Frankish Pepin le Bref retook Narbonne, and the province entered progressively into the kingdom of the Franks. With the domination of the Frankish kingdom and the advent of the Carolingian dynasty, we find ourselves in Transition 8.
13Transition 7 unfolds according to several time periods on several spatial scales. The historical frieze of figure 2, showing intersections between the different geographic levels on which power was exercised (lines) with the periods (columns) is a way of illustrating the imbrication of political powers (colours). It permits specification of the spatio-temporal contexts in which the process took place:
The new coverage of territorial power, with the coexistence of the ancient spatial organisation (diocese of the Gauls – the Seven Provinces – provinces, cities) and the Germanic kingdoms (Visigothic and Frankish for the space under consideration).
The institutional implantation and territorial coverage of the ecclesiastic dioceses, beginning in the fourth century.
The resilience of an inherited geography of power, which confers on certain capitals a supremacy that remains active, on both the political and religious levels, on the interprovincial, the provincial and the civic levels (respectively, Arles, Narbonne, city capitals).
The process of dismembering the ancient cities occasioned by the emergence of the castra, the multiplication of episcopal sees, the progressive installation of representatives of royal power, Visigothic and Frankish (counts, dukes), whose functional mobility makes use of the city capitals but also encourages the appearance of new political, judicial and fiscal places (such as the seats of counts in the Carolingian epoch: Substantion, near Montpellier, Ruscino, a degraded former Latin colony, Conflent and Redae-Razès, in Roussillon).
Figure 2: Spatio-temporal timeline of the transition: evolution of the imbrications of political powers.

Civic and ecclesiastical space: granularities and evolution
14The problem at issue is that of the structuration of the civic space, which in the beginning theoretically coincides – the point is debatable – with the space over which the bishop exercises his power.
The heritage of the Early Empire (third century)
15On the administrative and fiscal levels, the ancient civic framework (Regime 1) was subdivided into pagi, endowed with institutions and magistrates. Le role of the vici, a term designating agglomerations, both villages and towns, is still a subject of debate. To these locales were added the domains of the city notables and other property-owners which constituted the basic entities of property taxation (fundi)11. All these embedded territorial entities were relay points in the administrative and fiscal hierarchy12.
16On the level of settlement, the population was divided between dispersed habitat (farms, villae) and grouped habitat (hamlets, villages, small villages, small towns).
Evolution during late Antiquity
17From the third century to the mid-fifth, if the rural geography of the Early Empire persisted in the organisation of settlements, the habitat evolved, with continued reduction of the number of secondary agglomerations and concentration of settlements around centres which, while less numerous, were larger and larger, benefiting from new monumental construction. In the fifth and sixth centuries, a new rural geography was put in place, with 1) the creation of hill top and fortified agglomerations (castra) and 2) the appearance of a new stratum of establishments, where new architectures techniques were tried out (pit-huts, greater use of perishable materials) and which would constitute new population centres in the course of the sixth century (figure 3). These architectural innovations coexisted with the residences of the elite in stone, marble, roman concrete and tile roofs13, and reveal the individualisation (household plot) of a social category (dependants and small-scale peasants), the archaeological evidence of which is not really identified in Antiquity14.
18The canonical Merovingian texts distinguish the vici, the castra and the villae. Archaeology over the past three decades has revealed the creation, in the fifth and sixth centuries, of more than twenty hill top and fortified agglomerations (oppida, castra, castella) within the framework of the Septimania – whether created from scratch, by restructuration of an ancient agglomeration, or by re-occupation of a protohistoric oppidum15. This new generation of grouped settlements partly compensates for the abandonment or withering away of agglomerations enumerated since the second century. And the Liber judicum which compiles the laws of the Visigoths suggests ‘que le castrum tardo-antique disposait d’attributs administratifs et pouvait fonctionner comme un lieu d’exercice du pouvoir judiciaire, sinon comme une véritable circonscription administrative’ (that the castrum of Late Antiquity possessed administrative attributes and could function as a place of the exercise of judicial power, if not as a true administrative circumscription)16.
19In the former city of Nîmes, now subdivided into three, even four cities – Nîmes, Uzès, Maguelone et Arisitum (figure 2) – the villae continued to function as population centres, even if their functional structure had been thoroughly reformed and their monumental decoration diminished: the study of Loïc Buffat on the city of Nîmes still counts, in the sixth century, 37% of the 198 villae occupied during the Early Empire17, and of the 89 villae excavated pr prospected that were created or occupied in the first century B.C.E., 73% were still occupied in the fifth century, 42 % in the sixth and 25% in the seventh.18
Figure 3. Dynamic of the sites of power in the former territory of the ancient city of Nîmes in the sixth century: the network of agglomerations (including the castra) and villae, and the principal bishoprics19.

The ecclesiastic geography
20On the level of ecclesiastic geography, the arrangement confirmed in the Carolingian epoch was progressively put in place. The canons of the councils20 of the fifth and sixth centuries mention the territory of the bishop, territorium episcopi or territorium civitatis – what is termed a ‘diocese’21 – and the parochia, ‘parish’, which then designated the church in which services and the great liturgical feasts were celebrated, and respecting which one distinguished the churches intra muros of the episcopal see and the rural churches. The parochia, also known as diocesis, ‘diocese’, did not yet refer to a territorial reality. The Merovingian councils designated by the term parochia, distinct from the private oratory (oratorium), the churches constructed by bishops in the vici but sometimes also those which had been founded by great property-owners in the villae, when these had the right to baptise (cf. council d'Orléans de 541, canons 26 et 33). The bishop therefore had to accommodate the initiative of the great property owners and the monasteries in creating rural parish churches.
21The council of Tours in 567 mentions vicus archpriests – archpriests are also recorded in the castra – attesting to the installation of parish churches in the vici, which should undoubtedly be considered in part as rural agglomerations of Gallo-Roman origin22, but which also bear witness to the real capacity of Late Antiquity to create rural agglomerations23.
22The archaeology shows, in Septimania throughout the fifth and sixth centuries, a significant tendency to create churches and liturgical, baptismal and funerary complexes, identifiable in the castra and villae or in their proximity. A second wave of construction of churches and rural chapels in the seventh to eighth centuries would accompany and solidify the wider and deeper diffusion of Christianity in the countryside24.
The process of fragmentation of the ancient cities
The creation of bishoprics
23The process began in the fifth century with the emergence of bishoprics which no longer coincided with the framework of the city25 26. It should be stated at the outset that, if the sees of the bishops are known, their territorial substance is not. These bishoprics entered into competition with the power of the cities.
24Thus the city of Nîmes suffered from the secession of Uzès, which received, before 442, an episcopal see, and which was recognised as a castrum, perhaps as city capital, and of Ugernum-Beaucaire, a castrum attached to the city of Arles, which was still an administrative centre of imperial power in the fifth century. The city of Béziers lost the territory of Agde, which was endowed with an episcopal see, where the council of 506 was held. Around 540, the city of Nîmes was deprived of the mountainous zone situated between the Causse du Larzac and the Cévennes, which was elevated to a bishopric on the initiative of the Frankish kingdom: the bishop Mondéric was given the charge of fifteen ‘parishes’ there and ruled from the vicus of Arisitum, undoubtedly a castrum. Nîmes also lost Maguelone, which imposed its dominance as a port-city over the fourth and fifth centuries (figure 3).
25Further to the west, the city of Narbonne suffered amputations in several years. A Castrum in the fourth century, Elne received an episcopal see, whose incumbent is mentioned in 572: Narbonne lost is authority over Roussillon and the eastern Pyrenees. To the west, on the frontier of the Frankish kingdom, it lost the territory of Carcassonne, which received an episcopal see from at least 589.
The geography of royal power
26The Germanic kingdoms adjusted themselves to the geography of power inherited from Antiquity (figure 2, ‘Global 1, 2 et 3’). The traditional capitals took in the representatives of the Visigothic and Frankish power: Narbonne, capital of Narbonnensis prima, Arles, capital of the prefecture of the praetorium of the Gauls.
27Local power was entrusted to the counts, placed at the head of a circumscription containing a number of pagi and attested from the fourth century. The count was charged with levying soldiers for the royal army and presiding over the court of the county. A number are mentioned in Septimania (Agde, Narbonne).
28Prior to the Carolingian epoch, it is impossible, given the state of the sources available, to perceive geographic homogeneity and chronological continuity in the representation of royal power, Visigothic, then Frankish, in Septimania. It is in the Carolingian epoch, in the ninth century, that the Frankish power progressively organised more regular coverage by count-administered authority, based on former civic capitals and seats of the ecclesiastic diocese (Narbonne, Béziers, Lodève, Nîmes), on more recent civic and episcopal capitals (Uzès and Carcassonne), or on the episcopal capitals (Maguelone, Agde) (sixth and seventh centuries).
The factors in the disorganisation of the ancient civic framework
29Several sets of factors served to call into question the territorial frameworks inherited from Antiquity under Diocletian and Constantine:
Territorial competition between the Frankish and Visigothic kingdoms: this competition had several phases, as the Frankish conquest progressed. Defeated by Clovis at Vouillé, the Visigoths were compelled to yield Aquitania prima and fall back to their Spanish bastion, retaining, in Gaul, Septimania. The latter was the object of solicitations and territorial confiscations, as illustrated, in the 530s, by the Frankish creation of the bishopric of Arisitum on the north-western frontier of the city of Nîmes and by the conquest by Theodebert, the future Frankish king, of Uzès and Lodève27.
Competition between the regional high ecclesiastic authorities, the archbishops of Arles and Narbonne: creation of bishoprics and nomination of bishops took place in the context of a struggle for influence between the archbishops of Narbonne and Arles, each of whom claimed primacy over the bishoprics of Septimania and tried to impose his respective candidates, as at Lodève (possible initiative of the bishop of Arles in 421-422, opposed by the cleric of the city) and Béziers (before 461, nomination as bishop of the deacon Hermes by the bishop of Narbonne, contested by the clergy of Béziers).
Economic competition: the sixth century saw the growth in importance of ports in the new geography of power, with the bishoprics of Maguelone (figure 3), Agde and Collioure, in addition to the site of Narbonne. This promotion of port-cities illustrates the economic dynamism of the Languedoc coast and the high level of Mediterranean trade28.
The aristocracy and the territorial dynamic: the modifications that affected the territorial organisation of the province of Narbonnensis prima, then of the Gallic part of the Visigothic kingdom, raise the question of the general role of the Gallo-Roman aristocracy in the territorial ambitions of the ancient centres and the emergence of new poles of power, political and military, as well as ecclesiastic. The sources do not clearly convey the context of the creation of a new village or urban pole, and many establishments where aristocrats resided are ignored in the texts. But what is indisputable is the multiplication of places of power, which attests at once to the territorial competition between the political and ecclesiastic authorities on the regional scale (Visigothic and Frankish kingdoms, primacies29 of Arles and Narbonne kingdoms) and to the territorial competition on the meso- or micro-regional scale.
30A number of events testify to the capacity of the provincial elites to contest royal or primatial authority.
31The most serious event was certainly the attempted secession of the province in 672-673, when two Visigothic dignitaries, Ilderic, count of the city of Nîmes, and his ally Gulilde, bishop of Maguelone, revolted against the recently elected king, Wamba, and contested his legitimacy.
32Another domain in which local initiative was expressed was the contestation of decisions of the archbishops of Arles or Narbonne over the nomination of bishops (around 421-422, in Lodève and, shortly before 461, in Béziers).
33Another level of observation also offers means of measuring aristocratic initiative: that of the creation of castra, in the fifth to sixth centuries, attested at once in the texts, by the sudden appearance of these settlements in the official lists of cities, and the possible installation in them of an episcopal see, and by archaeological evidence, with explicit indications of aristocratic residence in certain cases. If some castra reactivated former ancient agglomerations (Ucetia-Uzès, Ugernum-Beaucaire, Carcasone-Carcassonne, Caesar’s camps at Laudun, Suzon, etc.), most were new creations, such as Le Roc de Pampelune, Saint-Pierre-de-Valliguières, Mormellicum, Saint-Julien at Anduze, etc. They established centres in spaces either endowed with an economic and social dynamism proven since the Early Empire, such as the north-east of the former city of Nîmes, between the Gardon and the Cèze, opening towards the middle and lower valley of the Rhone, or little settled in Antiquity, such as the hilly and mountainous sectors of the western part of the city30. These hill top and fortified settlements could not all pretend to the same status, nor to the same functions, as is evident from their structures, their construction and their size, which varies from less than 50 ares to more than 5 hectares. But the emblematic example of the castellum of La Malène, in Lozère, on the frontier between the Visigothic and Frankish kingdoms, which was occupied from the end of the fifth century to the end of the seventh, proves that this type of establishment could contain an aristocratic residence, strongly fortified and endowed with elements of decoration and comfort, including thermal baths, which added to the pleasure of living in it31.
34The scattered position of the castra then poses the question of the authority capable, on the local level, of bringing together and mobilising the population and the notables to erect places of strength and fortified small towns. Moreover, certain castra could accommodate an episcopal see, such as Uzès, prior to 442, Carcassonne and Elne in the second half of the sixth century, Collioure in the seventh century. It is clear that the implanting of the castra reflects the emergence of powers on the micro-regional scale, and that it is a means of tightening the territorial mesh observable at the level of the ecclesiastic map under construction during Late Antiquity.
35At the end of this transition, the geography of powers inherited from Antiquity has been thoroughly rearranged: to the system concentrated on the ancient city has succeeded a multiplicity of seats of power, distributed between the politico-military and the ecclesiastic hierarchies. The local aristocracy, the representatives of royal power and the bishops give rise to a new multipolar landscape in the process of stabilisation. This is the structure on which the Carolingian innovations will be based for determining a new geography of power.
From the narrative to its formalisation
36The point of view given by the structured narrative will serve as a basis for the formal description of the transition – that is, the functioning, hierarchic or not, of the different powers in play, their ways of interacting and their spatial implications. The transition corresponds essentially to rivalries between powers operating at different geographical levels and having implications for the organisation of space. In particular, changes at the intermediate level (city, province) will have an effect on the properties of entities at the local level, which, in turn, will have an impact on the functioning of the hierarchy. To begin with, we propose a representation of the transition by means of a sequence of schematisations of the functioning of these different powers at different stages. This will be followed by a description of the conceptual framework of the model, which can serve as a specifications grid in anticipation of computer implementation.
Hierarchical and spatial structures associated with the exercise of power
37In the spatio-temporal framework constituted by Septimania between the fourth and eighth centuries, the historical narrative makes it possible to relate the pressure of Germanic peoples to exogenous disturbances entailing a fragmentation and weakening of imperial Roman power. One observes an endogenous process of affirmation of ecclesiastic power. The process, previously noted, of diminishing the number of agglomerations which took place beginning in the second century will be considered as operating independently of the religious factor and of rivalries between powers. It will not be the object of formalisations, and can be integrated subsequently as a process ‘exogenous’ to the modelled system.
38The first step concerns the schematisation which can be established of the different powers and their functioning, hierarchical and spatial. We propose several representations, each corresponding to a ‘state’ of the political system. The first diagram (figure 4a) illustrates the functioning of Roman power and the territorial hierarchy associated with its exercise. With each of these territories, whatever the level, is associated a capital. It is the agglomeration of the territory where power is exercised, that is, where those persons who exercise it reside. The whole of the population residing in the territory concerned is therefore subject to this power: taxation, administration, justice, security… This population is distributed between dispersed and grouped settlements. With respect to this power, there is complete identity between the territorial hierarchy and the hierarchy of places where these powers are exercised, since one is dealing with the capitals of these territories (prefecture of the Gauls, province, diocese, city).
39This control is schematised by the links indicating the relations between the local power and the population residing in the territory: administration, taxation, justice, security… (figure 4b): it corresponds to the initial state of the political and territorial system that we have called regime 1 (R1), which corresponds to the third century.
Figure 4: Hierarchic (a) and spatial (b) structure of the exercise of power under the ancient Roman empire in regime 1 (R1).

40On the social level, power is a function of a single class of the population: the aristocrats. The existence of such a class in the agglomerations is that one among the elements that seems to be a necessary condition of – if it is not sufficient for – the presence of representatives of power in a place. The local aristocracy, Gallo-Roman, then Germano-Gallo-Roman, plays an important role and must be integrated into the formalisation. It constitutes a linking factor between the potentially competing forces represented by these different powers. It is an important agent of power. It is from this class that the different elites emanate; within it the different powers interact.
41The dynamic of the political system is founded on the forms of co-presence of the powers: the first step of the dynamic is illustrated (i.e. the initiation of the transition, which we shall call R1+ and which corresponds to the fourth century). This concerns the introduction of the functioning of the Germanic powers (Visigothic, Frankish), which is organised according to one identical structure. The exercise of the Germanic powers is related, from this point of view, to cooperation with the imperial power: Visigoths, later Franks, maintain the structures in place and retain the territorial frameworks of the Roman Empire. Figure 5a illustrates the fact that the Germanic powers integrate the Gallo-Roman aristocratic structure (alliances, marriages, ethnic mixture) and the functioning of the system of Antiquity. The Germanic and Gallo-Roman elites share the powers in Septimania. The space of political power is transformed without any impact on its spatial inscription and therefore on the geographic space, apart from the phenomenon of clarifying the web which, as we have said, relates to another process (figure 5b). However, if the structures of the different Germanic powers can be identified, their localisations differ: the Germanic powers do not inhabit the same space. Thus, Septimania is spatially divided between these different powers, Visigoth and Frank.
Figure 5: Hierarchic (a) and spatial (b) structure of the co-existence between the exercise of imperial power and that of the Germanic kingdoms during the period of the initiation of the transition (R1+).

42Can one similarly describe the ecclesiastic power over the course of the transition? What differences are there between these two forms of power, political and religious? At the beginning of the period, the religious element represented by Christianity cannot be associated with any power. It is a latent existing phenomenon which becomes active with the arrival of Constantine and the end of the persecutions, unlike the installation of the Germanic peoples. We therefore witness a progressive affirmation of these new religious power, which can be described by way of both analogy and difference to the preceding logic.
43The emergence of ecclesiastic power comes through the bishops, who, at the very beginning, are installed where political power exists at the most local level, that of the city and the city capitals. There is a recognition of mutual interest in the concentration of power in the same place. The aristocracy plays a pivotal role, with the bishops issuing from that class. Ecclesiastic power is therefore initially asserted within the same social and territorial framework as civil power. If the bishop has power over a bishopric or diocese whose territorial framework corresponds at the beginning to that of the city, the ecclesiastic hierarchy functions in a particular way, namely by constructing itself upwards from the bottom, and its territorial inscription does not follow this hierarchy, because the archbishop, who has an influence on the other bishops in the same province, governs a diocese as the other bishops do.
44The structure of the organisation of ecclesiastic power, if it is aligned with the power structure inherited from the Empire and assimilated by the royal powers (hierarchy of powers and territories associated), is therefore less well described in terms of a hierarchy and the associated embedding of territories (figure 6a). Ecclesiastic power is capable of spatial innovation, for it has neither attachments nor constraints in relation to the Gallo-Roman heritage (such is the case with Maguelone, Arisitum, Collioure, etc.). Figure 6b illustrates those displacements which appear in certain places as reflecting assertions of the exercise of power. Thus the fact that the superior echelon does not have the same basis allows for spatial rivalries and innovations with the emergence of new bishoprics, which contributes to dissociation and desolidarisation of the civic space by creating new capitals functioning on the political, ideological, economic, fiscal and even military levels.
45One can identify the principal components favouring the implantation of these innovations in a place:
the potential of the aristocracy, which is represented both in the political and the religious spheres;
the economic potential (see, for instance, ports such as Maguelone, Agde, Collioure, etc.);
the situation of the place in relation to the civic space: marginal and difficult-to-control places, less subject to the exercise of other powers, will be able to receive an installation (Arisitum, towards Le Vigan at the foot of the Causses).
46These characteristics are important in evaluating the probability of the emergence of a bishopric in a particular place. If the properties of places seem to be the most decisive factors, it appears that other factors should be taken into account:
at the regional level, the location in relation to the influence of the seats of external powers;
at the local level, the potential of churches, which represents, after the potential of the aristocracy, the second component of the establishment of religious power (e.g. Arisitum and its capital of 15 churches).
Figure 6: Hierarchical (a) and spatial (b) structure of the coexistence between the exercise of civil and ecclesiastic powers in the course of the transition.

47The conceptual framework of the model is formed around the processes of substitution of the Germanic powers for the imperial power, competition between the Germanic powers in southern France and diffusion of ecclesiastic power.
Proposition of a conceptual framework for modelling
48This initial reading of the transition constitutes the interface between the structured narrative and the formalisation of a model. It permits us to verify that there is a consensus on the identification of the mechanisms that characterise it and to move on to their more formal specifications, something which then allows for development of the ontology of the computer model32 33in order to test whether they are sufficient to account for the observed facts and mitigate our ignorance of the logics governing the decomposition of the territories of power. We propose now to specify the entities of the model, their properties and the relations among them.
The entities of the model
49Three broad types of entities have therefore been identified: two types of spatial entities and one type of social entities. The definition of the entities is stable over the transition. What evolve are the values of their properties and the relations that exist between them. We present each of these classes of entities one by one. Their properties are listed and described in the tables recapitulating their role and evolution, if appropriate. We will then discuss the processes or conditions which underlie their alterations.
50- The urban entities: this class is essentially composed of urban agglomerations which constitute the basic structure of occupation of the territory by populations. They are organised according to a spectrum of size: from the rural agglomerations (with few functions) to the agglomerations with urban functions. In the agglomerations, one finds the capital cities, which are the seat of civic and/or ecclesiastic power. The other agglomerations are under the political supervision of a capital city (city, province, regional prefecture of the praetorium). These entities are long-lasting; change does not transform them in an intrinsic way.
51Table 1 recapitulates the properties of the entities categorised according to three types:
the intrinsic properties which describe the entity as such: it is the aristocratic component, a vestige of the past status;
the extrinsic properties, which are acquired by way of external entities. These are the seats of one or several powers and the associated territory of influence;
the contextual properties, which are the properties evaluated by calculation in relation to the entirety of the system: they are relative localisations and accessibilities.
52Certain of these properties do not evolve over the course of the transition. For those that evolve, the rules and factors associated with the evolution according to the context are given.
53The aristocracy, which is central in the process of implantation of powers and their instabilities, represented here by the ‘aristocratic component’, is reduced to a score which varies partially with the ‘number and levels’ of the powers in place in the location, and partially in an endogenous manner. It thus constitutes a potential trigger of crisis, according to the ratio between the ‘demand’ (aristocratic component) and the ‘supply’ of power (powers in place in the location). If the ratio is high, this signifies that the aristocratic component is elevated with respect to the power in place and that the agglomeration will potentially be the seat of a crisis, which may provoke delocalisations of power.
54The urban agglomerations are the scene, on the one hand, of a rivalry among powers, but also a diffusion of powers. They receive the active entities of the model.
55- The rural entities: this category of entities includes two types of spatial entities which have been identified as markers in the emergence of places of power according to a ‘bottom up’ logic.
the agglomerated entities: the vici should be considered with regard to their capacity to receive churches and constitute themselves as ‘parishes’. A spatial concentration of these characteristics will constitute, in a particular area bordering an urban agglomeration, one of the conditions necessary for the appearance of bishoprics.
the dispersed entities: the large domains or villae are isolated rural entities – places of residence of the aristocratic society and of power (secular or religious). When they are inhabited, these places become attractive and potentially dynamic. They can then receive churches.
Table 1: Properties of the entity ‘Urban agglomeration’.

56As for the urban entities, table 2 synthesises their properties. Two intrinsic properties have been identified: the aristocratic component, which marks the social dynamic of the place (which will contribute, for example, to the construction of fortifications), and the site. The entity may or may not accommodate a church: this is an extrinsic property, because it is the result of a process which acts in the background (see above). Finally, the entities are described by a contextual property, namely, their relative situation with respect to the different sites of exercise of power.
57The entities in this category are passive. They receive the churches on which ‘bottom-up’ dynamics are based.
58- The local powers: this category is formalised as social entities. They are localised in the spatial entities and have a certain life-span. They symbolise the exercise of power and are therefore of a certain type (Visigothic, Frankish or ecclesiastic) and of a certain hierarchic level, and they depend on a certain power of a superior hierarchic level (‘power+1’) (figure 7). It is thus possible to re-compose the graph of hierarchic relations associated with a type of power, whatever the form, on several levels for the Germanic powers, on only two for the ecclesiastic powers. Moreover, if one maintains the hypothesis that the territories are attached to the closest local power, these graphs then permit re-composition of the embeddedness of territories.
Table 2: Properties of the rural entities.

59Figure 7 summarises the sketch offered here of the entities, properties and relations among the entities. Indicated there are the three classes of entities described above, with their properties. In this diagram, the spatial component has been added in the form of a class playing the role of integrator in the relation which proceeds from the rural entities to the power, and inversely. As was previously stated, the capital agglomerations control portions of territories. In the initial situation, the portions of territory are under the control of the nearest capital. These controls will evolve over time and be recalculated according to the redistributions of powers across the urban entities. The overall situation has been projected on the basis of a territory illustrating one province, composed of four cities (Narbonne, Béziers, Lodève, Nîmes) and forty or so agglomerations34. Over the territory about 250 rural entities are distributed (villae and vici). The initial situation will be fitted onto the situation observed, with the local powers distributed in the four city capitals.
Figure 7: Diagram of the entities, properties and relations.

Dynamic
60The dynamic of the system will have to be constructed around the question of the appearance or disappearance of a power in a place. It may be defined in relation to a number of mechanisms which are now presented according to two categories, depending whether they are exogenous or not. It will give rise to calculations of new attributes, which will have to be integrated into the diagram of figure 7.
61- The dynamic of political powers: The pressure of the conquering Germanic powers is a ‘triggering factor’ of the dynamic of reorganisation. It is an exogenous process which has its own dynamic. Each place is subjected to this exogenous pressure (figure 8).
Figure 8: Competition between the Germanic powers affecting the territory.

62This ‘pressure’ is applied to the places overall, but only certain places are ‘eligible’ because they fulfil the necessary and sufficient conditions for the localisation of a power, namely:
seat of a power (at the outset: geography of power inherited from Antiquity);
effect of a favourable site or situation;
sufficient separation from another place of competing power of a superior level;
sufficient level of aristocratic component.
63The totality of these combined conditions defines a score of ‘eligibility’ for the central power’s emission of a ‘local power’ entity to a place.
64The places, however, oppose to this pressure a stronger or weaker ‘resistance’, or else no resistance at all. This ‘resistance’ is a function of the coherence of the power within a place, that is:
existence of an established competing (Germanic) power and intensity of the pressure of the power on which it depends (figure 8) ;
intensity of the established power (number of powers and level of these powers);
level of the aristocracy and harmonious functioning of power (relation between the quantitative level of the aristocracy and the potential for exercising power).
65It is the ‘pressure-resistance’ confrontation in the ‘eligible’ places which will determine the probability that a power will install itself in a place. This installation may be effected either directly, if no ‘local power’ entity is already installed, or by substitution for a ‘local power’ already installed.
66- The dynamic of ecclesiastic power: In contrast with the civic power, whose dynamic is triggered by an exogenous process at the macro-scale level, the dynamic of ecclesiastic power is founded on a process at the micro-scale level, that is, on the evolution of the amount of churches and aristocracy of the domains and rural agglomerations. It is in this regard that one observes a certain number of spatial innovations.
67To start with, the conditions of the ‘appearance’ of a ‘local power’ entity of the ecclesiastic type in a place depend simply on the following conditions:
seat of a civic power,
sufficient level of aristocratic component.
68Subsequently, the mechanism of appearance may detach itself from the first condition (existence of a civic ‘local power’ entity) and benefit from an establishment linked to the potential of churches in the area and/or to the dynamic of the local aristocracy. These new conditions determine a score of ‘eligibility’ of urban agglomerations which it will also be necessary to confront with the pressures of civil power on the place. According to the result of this confrontation between ‘pressure of civil power’ and ‘eligibility of ecclesiastic power’, a ‘local power’ entity of the ecclesiastic type may appear in this place.
69The conceptual framework which we propose, then, is essentially founded on these two processes, the way in which they will interact and the types of spatial configurations which this interaction will cause to emerge.
70In light of these specifications, different questions take shape which might be asked concerning this representation of the phenomenon, especially the respective weight of the conditions of ‘eligibility’ of places to receive a power: what is the weight of the initial spatial configurations? What is that of the effects of site and situation? Is the issue a simple remeshing of the space with a finer granularity initiated by the appearance of a ‘bottom up’ logic, or is it rather the product of the interaction of this logic with that of the rivalry of hierarchic powers?
Discussion and perspectives
71The study we have just proposed finally corresponds to the first step of experimentation, that of thought, and to its expert validation. This thought experiment is very rich in itself: it calls for elucidation, especially by breaking down the empirical knowledge of the imbricated processes. The formal elucidation is founded on what appears, and the passage to a generic level reveals what is missing. The validation comes from the ‘authorisation’ by the expert of the passage to this level.
72The step of defining the conceptual framework of the model has made it possible to specify the point of view adopted to represent the transition – that of the functioning of powers – and to identify the generic entities, while integrating the historic specificity. If we have not completely defined an operational syntax, we have placed the markers for it. This framework was conceived instead in the perspective of the development of a multi-agent system. The definition of the ontology of the model will permit resolution of various ambiguities and specification of the entire group of functions and operations associated with different mechanisms. Given the point of view adopted, one could have formalised the question in a framework derived from game theory, and explicitly studied the question of rivalry or cooperation between powers, and between powers and aristocracy35 36. This question is not closed, and it may provide inspiration for the formalisation of the interactions between powers, and between powers and aristocracy, on the local level, that of capitals on the micro-regional scale. The idea would be to delegate, within a multi-agent system, these mechanisms on the local level, something that would permit integration of the spatial dimension in the mechanisms of competition and substitution (Germanic powers) or of co-existence and accommodation (Germanic power–ecclesiastic power) on the local level. The following step will therefore consist in conceiving the operational form of the model to explore the roles of the different factors identified with respect to the spatial configurations.
73The link between the empirical domain and that of the model is in place37. This step has made it possible to open up thematic questions: in the first place, that of the territorial substance of the spatial circumscriptions – cities, bishoprics, counties, duchies… Secondly, that of aristocratic initiative, its territorial inscription, its economic foundations, its operational localisation in acting on the mechanisms of decision, on both the political and religious levels. Moreover, from the methodological point of view, among the questions this step recalls, that which seems to us most compelling and constraining for this exercise is that of the spatio-temporal attachment of history. When the objective is to study and account for temporal factors and for organisations and rhythms in space, how to move from the time of the narrative to the time of the model (linear, continuous)? How to integrate within generic rules all the specificities of the historical dependences (‘path dependence’) of the spatial entities considered? If the experiment is unique from the thematic point of view, it is not so from the methodological one, and it must be possible to deal transversally with the questions it raises by integrating the different experiments performed in the TransMonDyn project – something which is proposed in the second part of the volume.
Notes de bas de page
1 The civitas, is the territorial entity constituting the first level of political organisation of the Roman empire: a civic community, its institutions (assembly, local senate, magistrates), a civic territory.
2 This position is simply due to the specialisation in Antiquity of the thematician who is co-author of this text. It was natural to start with a state that he knows relatively well, rather than define the final state in a highly complex period far removed from his experience and expertise...
3 Raynaud Claude, ‘Les campagnes rhodaniennes : quelle crise’ dans Le IIIe siècle en Gaule Narbonnaise. Données régionales sur la crise de l’Empire, Aix-en-Provence 1995, ed. J.-L. Fiches, Sophia Antipolis, APDCA, 1996, pp. 189-212.
4 Raynaud Claude, ‘Les campagnes languedociennes aux IVe et Ve siècles’, in Les campagnes de la Gaule à la fin de l’Antiquité, ed. P. Ouzoulias, Ch. Pellecuer, Cl. Raynaud et al., Antibes, APDCA, 2001, pp. 247-274; ‘De la conquête romaine au Moyen Âge’, in Les agglomérations gallo-romaines en Languedoc-Roussillon, Lattes, ed. J.-L Fiches, ADAL, coll. Monographies d’Archéologie Méditerranéenne 13 and 14, 2002, pp. 39-53.
5 The Tetrarchy initiated the partition of military power. A century later, at the death of Theodosius in 395, the empire was divided into two vast territorial conglomerates: the empire of the West, with Rome (or Ravenna) as its centre, and that of the East, centred on Constantinople.
6 The term ‘civic’ is used for what relates to the city, which constitutes the basic spatial entity of the provinces and of the Roman empire: a population (an electoral tribe), a territory, a capital, institutions (assembly, local senate or curia, magistrates, priests), powers of taxation.
7 Septimania appears in a letter of Apollinaris Sidonius dated 472. The word, on the model of Novempopulania (the province with nine peoples), might designate the province with seven bishoprics (Elne, Narbonne, Béziers, Agde, Lodève, Maguelone et Nîmes).
8 Traditionally claimed by specialists in ancient Rome and by medievalists, Late Antiquity now occupies a specific place, with an identity affirmed by a specialised historiography and archaeology. To it may be assigned, for Gaul, the Late Empire of the fourth century, the Germanic kingdoms from the fifth centurry on, and the Frankish Merovingian dynasty (457-751).
9 The Roman colonies were founded as such, generally installed in indigenous agglomerations and peopled by Roman colonists, civilian or military: in the case of Narbonne and Béziers, these were veterans of Caesar’s legions, the tenth and the twelfth, thus soldiers who had honourably completed their active service. The Latin colonies were peregrine indigenous cities – hence Gallic – that had received Latin law, which granted the civil rights of Romans to indigenous inhabitants and to their aristocratic elite the possibility of becoming Roman after exercising as magistrates in the city. These are honorary colonies.
10 Schneider Laurent, ‘Aux marges méditerranéennes de la Gaule mérovingienne. Les cadres politiques et ecclésiastiques de l’ancienne Narbonnaise 1ère entre Antiquité et Moyen Âge (Ve-IXe siècles)’, in L’espace du diocèse. Genèse d’un territoire dans l’Occident médiéval (Ve-XIIIe siècles), ed. Fl. Mazel, Rennes, PUR, 2008, pp. 69-95.
11 Forma censualis, according to Ulpien, dated 211-217: Dig., 50, 15, 4.
12 Tarpin Michel, ‘Organisation politique et administrative des cités d’Europe occidentale sous l’empire’, Rome et l’Occident, PALLAS, 80, 2009, pp. 127-145.
13 Schneider Laurent, Recherches d’archéologie médiévale en France méditerranéenne Formes et réseaux de l’habitat, lieux de pouvoir, territoires et castra du haut Moyen Âge languedocien (VIe–XIIe siècle), mémoire d’HDR, Université François Rabelais, Tours, 8 février 2013.
14 Zadora Rio Élisabeth, ‘Early medieval villages and estate centres in France (c. 300-1100)’, in The archaeology of early medieval villages in Europe, ed. J. A. Quiros Castillo, Documentos, 2009, pp. 77-98.
15 Schneider Laurent, ‘Structures du peuplement et formes de l’habitat dans les campagnes du Sud-Est de la France de l’Antiquité au Moyen Âge (IVe-VIIIe s.). Essai de synthèse’, Gallia, 64, 2007, pp. 11-56, pp. 34-38.
16 Schneider L., ‘Structures du peuplement…’, op.cit, p. 37.
17 Buffat Loïc, L’économie domaniale en Gaule Narbonnaise, Lattes, UMR 5140, coll. MAM, 29, 2011.
18 Pellecuer Christophe, Pomarèdes Hervé, ‘Crise, survie ou adaptation de la villa romaine en Narbonnaise Première ? Contribution des récentes recherches de terrain en Languedoc-Roussillon’, in Les Campagnes de la Gaule à la fin de l'Antiquité, ed. P. Ouzoulias, Ch. Pellecuer, Cl. Raynaud et al., Antibes, APDCA, 2001, p. 503-532 (4th conference of the association AGER, Montpellier, 11-14 March 1998).
19 Schneider Laurent. Recherches d’archéologie médiévale en France méditerranéenne Formes et réseaux de l’habitat, lieux de pouvoir, territoires et castra du haut Moyen Âge languedocien (VIe–XIIe siècle), Université François Rabelais, Tours, 8 février 2013.
20 The concile is an assembly of bishops whose decisions are formalised as canons, or laws. A distinction is made between general and regional councils, certain of which were held in Septimania: Béziers, 356; Nîmes, 396; Agde, 506.
21 Delaplace Christine., ‘Les origines des églises rurales (Ve-VIe s.). A propos d’une formule de Grégoire de Tours’, Histoire et Sociétés rurales, 18(2), 2002, pp. 11-40; Mériaux Charles, ‘De la cité antique au diocèse médiéval. Quelques observations sur la géographie ecclésiastique du nord de la Gaule mérovingienne’, Revue du Nord, 3, 2003, pp. 595-609.
22 Delaplace Ch., ‘Les origines des églises …’, op. cit., p. 23.
23 Zadora Rio E., Des paroisses de Touraine…, op. cit.; Schneider L., ‘Structures du peuplement …’, op. cit. (with regard to the castra).
24 Schneider L., ‘Les églises rurales de la Gaule (Ve-VIIIe siècles). Les monuments, le lieu et l’habitat : des questions de topographie et d’espace’, in L’empreinte chrétienne en Gaule du IVe s au IXe siècle, ed. M. Gaillard, Brépols, Turnhout, 2014, pp. 419-468.
25 Raynaud Claude, ‘De l’archéologie à la géographie historique : le système de peuplement de l’Âge du Fer au Moyen Âge’, in Peuples et territoires en Gaule méditerranéenne. Hommage à Guy Barruol, ed. M. Bats, B. Dedet, P. Garmy et al., RAN, suppl. 35, Montpellier, 2003, pp. 323-354.
26 Schneider L., ‘Aux marges méditerranéennes…’, op. cit. pp. 69-95.
27 Lodève returned to the Septimania from the second half of the sixth century.
28 Schneider L., ‘Aux marges méditerranéennes…’, op. cit., fig. 1; Schneider Laurent, ‘Castra, vicariae et circonscriptions intermédiaires du haut Moyen Age méridional (IXe-Xe siècle): Le cas de la Septimanie-Gothie’, in Écritures de l’espace social. Mélanges d’histoire médiévale offerts à Monique Bourin, ed. D. Boisseuil, P. Chastang, L. Feller, J. Morsel, Paris, Publication de La Sorbonne, 2010, pp. 237-266, pp. 239-240.
29 The ‘primacy’ is the rank of a ‘primate’, a bishop who exercises supremacy, of a variable kind, over all the bishops and archbishops of a region. ‘Primatial’ refers to the authority and the cathedral church of the primate.
30 Schneider Laurent, ‘Cité, castrum et “pays” : espace et territoires en Gaule méditerranéenne durant le haut Moyen Âge. L’exemple de la cité de Nîmes et du pagus de Maguelone (Ve-XIe siècles)’, in Le château et la ville. Espace et réseaux, éd. par P. Cressier, Casa de Velazquez, Madrid, 2008, pp. 29-69, pp. 39-40.
31 Schneider Laurent, Clément Nicolas,’Le castellum de La Malène en Gévaudan. Un "rocher monument" du premier Moyen Age (VIe-VIIe s.)’, Carte Archéologique de la Gaule, La Lozère (48), Académie des Inscriptions et Belles Lettres, Comptoir des presses d'universités, 2012, pp. 317-328.
32 Phan Denis, ‘Ontologies et modélisation par SMA en SHS’, in Ontologies et modélisation par SMA en SHS, ed. D. Phan, Paris, Hermes, Lavoisier, 2014, pp. 53-94.
33 Tannier Cécile, Zadora-Rio Elisabeth, Leturcq Samuel, Rodier Xavier, Lorans Elisabeth, ‘Une ontologie pour décrire les transformations du système de peuplement européen entre 800 et 1100’, in Ontologies et modélisation par SMA en SHS, ed. D. Phan, Paris, Hermes, Lavoisier, 2014, pp. 289-310.
34 Schneider L., Recherches d’archéologie médiévale…, op.cit.
35 Hartshorn Max, Kaznatcheev Artem, Schultz Thomas, ‘The evolutionary Dominance of Ethnographic Cooperation’, Journal of Artificial Societies and Social Simulation, 16(3), 2013, p. 7.
36 McCalley James, Zhang Zhong, Vishwanathan Vijay, Honavar Vasant, ‘Multiagent negotiation models for power system applications’, in Autonomous Systems and Intelligent Agents in Power System Control and Operation, ed. Ch. Rehtanz, Springer, 2003.
37 Livet Pierre, Sanders Lena, ‘Le "test ontologique" : un outil de médiation pour la modélisation agent’, in Ontologies et modélisation par SMA en SHS, ed. D. Phan, Paris, Hermes, Lavoisier, 2014, pp. 95-110.
Auteurs
UMR 6249, CNRS-Université de Bourgogne-Franche-Comté, Besançon.
UMR 5600 EVS, CNRS-Université de Lyon, Lyon.
UMR 7298 Laboratoire d’Archéologie Moderne et Médiévale en Méditerranée, CNRS-Université de Provence, Aix-en-Provence.
UMR 5140 Archéologie des Sociétés Méditerranéennes, CNRS-Université de Montpellier III, Montpellier-Lattes.
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