Les caves de Reims/Durocortorum à travers les exemples de la fouille du tramway

Abstract

In 2007-2008, the excavation of sector 4-5 along the Reims (Marne) tramway route, between Place de la République and Square Colbert, led to the discovery of some thirty cellars, backfilled between the middle of the 1st c. and the end of the 3rd c., and even the beginning of the 4th c. AD. In this sector, more than 8,000 m2 of ancient remains are related to occupations along five streets and five blocks, cut off because of the linear nature of the development project. The tramway cellars are another in a long list of buried masonry structures, of simple or more complex construction style, sometimes lightly decorated, a common feature in the marginal neighborhoods of the Early Roman city. They are rarely preserved in their entirety within the excavated area due to its linearity, as well as the impact of modern development. They are either badly affected by recent disturbances, or located at the limit of the excavation area and very often merely a corner of their walls is present and documentable. We were therefore not able to determine the outline or total dimensions of each of the cellars, and it remains impossible to draft a complete typological list based on the available criteria. Taken as a whole, however, the data do provide a good overview of the modes of occupation present within this intermediate urban zone, located between the Porte de Mars and the banks of the Vesle. In this zone, modest and more affluent dwellings, artisanal practices, as well as, in all likelihood, commercial activities would have benefited from the proximity of the streets, mixing these different worlds together, in a relatively anarchic manner, as the limits of the different housing units are difficult to determine with certainty. When the standing remains are nearly non-existent, as they were in this case, the cellars’ backfill often proves to be invaluable for characterizing the associated occupations. The materials and construction techniques used in the cellars are listed and analyzed, as well as all the elements that might indicate uses other than the simple storage of foodstuffs. While cellar function is inevitably addressed thematically, it remains in most cases unresolved, for lack of convincing evidence. In this comparative approach, the data collected, whether material, spatial or chronological, are replaced in a broader context, be that on the scale of a city block, the site, more generally, or the ancient city as a whole. This in turn led us to return to certain discoveries that reopened the question of the religious function of these basements in Antiquity. The tramway cellars represent a characteristic sample of Gallo-Roman cellars in Reims, in terms of location, morphology and architectural features. The study of the spatial distribution of cellars within a city block demonstrates that the majority of cellars (40%) were located streetside. The average surface area of complete cellars reaches 11.90 m2. In accordance with what is observed elsewhere in Reims, the plans of the cellars are generally simple, typically rectangular, though some rare instances are square or more complex in layout. The construction techniques used are generally common and fairly standardized. The walls’ masonry is composed of gritstone (sometimes, though, rarely of limestone rubblestone) and/or of chalk rubble with, in some cases, terracotta used, especially in niches or along the top level. The decorative elements observed on the walls remain limited and are based, a priori, on the alternation of materials used (terracotta, chalk or limestone) in order to create a contrast of color, as well as on a particular treatment of the mortar joints (joints drawn with iron, sometimes enhanced with the addition of a red color). The use of rubble or chalk blocks in wall masonry is more frequent in the most recent cellars. Conversely, the walls built with tiles seem older. Apart from these exceptions, the hypothesis of an abandonment of more rudimentary storage structures in favor of cellars seems to be confirmed by the tramway archaeological data and this transition likely occurred around the middle of the 1st c. AD. More than half of the cellars of sector 4-5, were sealed off when the site was abandoned toward the end of the 3rd c. and beginning of the 4th c. Before that, the same number of cellars were backfilled at the end of each phase, between the end of the 1st c. and the middle of the 3rd c. If we consider the chronological phasing of the sector 4-5 remains, it would appear that the pace of development was the same for the cellars and the adjacent habitat. The construction boom within the residential blocks in the second half of the 1st c. AD, and then between the middle of the 2nd c. and the beginning of the 3rd c., never seems to coincide with the major works undertaken in public spaces during the first half of the 2nd c., but rather, always lags behind them. Completion of the construction program in the private sphere occurs during the first half of the 3rd c. AD, with the installation and generalization of hypocausts, concrete floors and mosaics (attested to here, in the fragmentary state). These observations demonstrate the need to systematically consider cellars within their context of discovery, and at all relevant scales of analysis: the site, the city block and the city as a whole, in Antiquity. The objective is to see if the pace of construction for the city as a whole are reflected in the cellar chronology. As far as the purpose of the cellars are concerned, several clues point to their multi-functionality. Contrary to other examples from Reims, none of the cellars from the tramway excavations are significantly different from other cellars in the area in terms of their internal fittings. Only a few objects are related to ritual or cult activities (e.g., a three-headed statuette) or were used for artisanal production, but none of the latter have been discovered in a context related to their function. Drawing correlations between the location of discovery, the type of activities encountered on site, the links with other vestiges, and the category of material culture lend sense to the analysis. Examination of the tramway cellars’ plan, dimensions and architectural data did not reveal any functional distinction. This analysis also represented an opportunity to remember that the Gallo-Roman cellars discoveries in Reims do not only concern the marginal zones, with an artisanal or commercial vocation, but also occur in the heart of the ancient city. Less than 500 m from the forum, the cellars within which bread rolls and ex-votos were found in the rue Saint-Symphorien, as well as the cellar within which the zoomorphic vase in rue de l’Écu was found, are among the most emblematic examples. The rue Saint-Symphorien site highlights the need to search for potential links between the cellars and the above ground rooms, all the more so in burnt-out contexts favorable to the conservation of organic materials remains. The second example, which also includes evidence of fire, demonstrates that even a shallow cellar, that has been particularly impacted by modern construction, can yield a large quantity of material culture, including objects that are original and rich in information.