Tempering with white marbles: implications on the provenance and production technology of ancient pottery

The addition of temper in the pottery production is attested since prehistoric times and is still a production choice adopted in modern ceramic industry. The type of temper comprises natural lithic materials (such as sand and rock fragments), materials of organic origin (vegetal or animal), as well as artificial products like grog/ chamotte (pottery fragments). When the temper is composed of minerals and rocks available in the area of the production site, they can be considered as useful markers to trace the provenance of pottery. On the contrary, the occurrence of different rock types outcropping in two or more geological areas far away each other, open new questions on the production technology. In fact, such situations can be explained only considering the recycling of imported rocks, including those used for architectonic elements or sculptures, mainly coming from buildings of previous epochs. This study presents the evidence of the deliberate addition of recycled white marbles on a long term production (4th and 14th century AD) of coarse and cooking ware in north-eastern Italy. The petrographic analysis of about 200 potsherds, attested the use of this unusual kind of temper, on about half of the repertoire. The occurrence of different type of marbles, associated to rocks and minerals typical of the alluvial deposits of the eastern Po plain as well to locally available rocks (Euganean Hills trachyte), clearly pointed out the intentional addition of recycled marble fragments, and allow to exclude that the pottery was imported from other regions. Detailed petrographic analysis on the microstructure, maximum grain size (MGS), accessory minerals (when observed) and grain boundary shapes, allowed to constrain to the most important Mediterranean classical sources area Aegean the provenance of the marbles. These data were also partially confirmed by the stable oxygen and carbon isotope analysis on marble fragments mechanically separated from the ceramic paste. Nevertheless, part of the marble fragments were characterised by very negative δ13C values, not consistent with any marble variety. A series of firing experiment on two marbles were carried out in the temperature interval between 600°C and 800°C, both reproducing oxidising and reducing conditions, to evaluate whether these anomalous δ13C values could be related to the firing process.