Version classiqueVersion mobile

Mottes castrales en Provence

Daniel Mouton


Texte intégral

1This work presents nearly two decades of research on the first stages of the castle phenomenon, at the moment of the militarization of territories that developed in Provence and numerous other regions around 1000 AD. At this time, numerous small terroirs developed under the domination of fortifications that are sometimes called “Moutte”, the Provencal version of “motte”.

2The zone studied is concentrated around the central part of the Provence region (fig. 1), principally in the Alpes-de-Haute-Provence and Bouches-du-Rhone Departments. Outside the coastal zone, the population density was never very high, thus allowing the conservation of numerous sites. Most are now found in forest undergrowth, however, making prospection rather difficult. A few sites were found by sight alone, but in many cases we were aided by texts describing a castrum. The charters of the Saint-Victor de Marseille and Cluny abbeys were very useful in this sense. Toponymy is also an effective method, but produces a great variety of sites since these small establishments are designated by toponyms as divers as motte (mound), castelas (castle) and other derivations of castel, roque (rock), tour (tower), or even puy (podium). Sometimes a site is designated as both a motte and a roque. On the other hand, denominations based on the term castel can be associated with mottes, but also with more recent castles or Protohistoric fortifications, or even natural stone formations. Numerous sites remain to be discovered and our research continues.

3The excavation of the site of Niozelles (Alpes-de-Haute-Provence), designated in texts from the beginning of the 11th century as roca de Aldefred, was very rich. This motte was installed on a natural formation whose summit was flattened to accommodate the first construction. This was a common practice in Provence where we currently know of only one entirely artificial mound. Around 970-980, a large rectangular house was built from local stone and a mud mortar (fig. 4, 5, 6 and fig. C, D, f.-p. pl.). This structure had no fortification; its only defensive element was its perched position on the mound. In one corner, a small aedicule (shrine) probably contained a staircase. We were able to confirm that this building had two floors based on the excavation of another site, La Moutte à Valensole. The Niozelles house opened onto a small court in which we found two wooden domestic installations: a small shed and a structure on four large stakes that can be interpreted as a granary or a place to stock wheat sheaves. This establishment was probably not very different from some contemporary houses and its construction required only vernacular techniques and modest resources. This first occupation lasted only around ten years. Around 980-990, the site underwent a radical transformation, becoming a true fortress, and thus now requiring significant resources and the skills of specialists. The first edifice was demolished and the next phase of construction was immediately undertaken. A large amount of earth and stones were added to enlarge the platform, in the center of which a rectangular donjon was built. The construction required the use of lime mortar, whose mixing pits we found. After a ritual deposit was made in the foundation (a coin in a pot), the tower was erected. It measured 9.25 x 6.75 m, with a basal width of 1.72 m for the walls (fig. 9, 10 and fig. E, F, f.-p. pl.). Once finished, this edifice was surrounded by a polygonal rampart also made from stones and lime mortar, protecting a space of 21 to 22 m diameter. Between the donjon and the rampart, all of the available space was occupied by domestic sheds, two made of wood and one of stone. Two of them protected hearths installed on the ground. Around 1030, the site lost its function as a castle. Later, an oil press was installed and probably operated for a few years before the site was definitively abandoned. We also found a large number of objects at this site. The coins allowed rather precise dating of the site. The ceramic artifacts constitute a catalog of forms that contribute significantly to our knowledge of daily life in Provence around 1000 AD. The metal objects revealed a series of domestic activities. In the category of weapons, only one arrowhead and some crossbow bolt heads were found. The study of the very abundant faunal remains provided important information concerning the meat products consumed, animal husbandry activities, and certain culinary practices.

4The motte of Niozelles is part of a territory in which we observe two occupation centers during the late Antiquity period. One is near the site and the other in area around the Medieval castle, which is now in ruins. At the location of this latter site, we cannot say whether the first occupation was contemporary with the motte or if the motte was installed after its abandonment. The house, which preceded the fortification on the motte, could have been built there by accident and have no relation with the castle site. But this is not the case since this type of construction has been found on mottes in Provence.

5This work continues with the study of more than thirty sites at which common characteristics and differences in organization have been recorded. In addition, when possible, the study of texts allowed a chronological approach to the ensemble of sites. In terms of morphology, the great majority of these sites consist of constructions on natural reliefs. In more than 70 % of the cases, these are stone knobs protected by a ditch that nearly always measures at least ten meters wide. Along with this defense element, we sometimes find the traces of a wall, and more rarely, the remains of a tower that can be rectangular, square (fig. 41, 42), or even round in one case (fig. 96, 97).

6The chronology revealed by the texts and excavations shows that these sites appeared in the last decades of the 10th century. Some of them were occupied until the 14th century and were probably subjected to many transformations. Nevertheless, according to current knowledge, the most recently constructed motte is that of Cadarache (fig. 71), whose occupation at the beginning of the 13th century was totally ephemeral.

7Two other sites are particularly remarkable: La Moutte à Valensole (fig. 48, 49) and La Moutte à Allemagne-en-Provence (fig. 52, 53). In both cases, the edifice is similar to the house revealed in the first phase at Niozelles, which was thus not a unique case. We can conclude that the militarization of these terroirs developed rapidly and progressively starting in the second half of the 10th century. The true fortresses appeared later, starting around 1000 AD. Others were built at the beginning of the 13th century.

8In some cases, we were able to observe the relationships between the constitution of a castrum and the occupation of a territory. In the lower valley of the Durance, for example, around a dozen sites are separated from each other by approximately 3,500 meters, thus indicating a high density. At Valensole, the texts of the Abbey of Cluny show that the territory of La Moutte attained 16 km2 at the beginning of the 11th century. At Allemagne-en-Provence, three mottes exist along with the Medieval castle site. We observe movements that occurred as early as 1000 AD with the destruction of establishments, followed by reconstruction on other sites.

9This ongoing research shows that the early stages of private fortification was integrated in a vast trend of the occupation of terroirs surrounding complex fortifications.

© Éditions de la Maison des sciences de l’homme, 2008

Conditions d’utilisation :

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search