First preliminary remarks on the petrography of the Dukki Gel ceramic samples

In the last few weeks I haven’t been very present in our blog since I spent much time sitting at the microscope of the Department of Earth and Environmental Sciences of the LMU, just nearby to our office, examining and documenting the first batch of ceramic samples from the site of Dukki Gel. These samples have been included as a reference collection within our DiverseNile project thanks to the kind agreement of the excavator, Charles Bonnet, and the responsible ceramicist Philippe Ruffieux. Philippe has already studied all of these samples within their context and we can now address fresh questions within the DiverseNile work packages and with scientific analysis.

In times of the Covid pandemic, the procedure to access the laboratories is rightly strict: registration is mandatory before working in the microscopy room, only a maximum of three people are allowed to work simultaneously in the lab and of course we are required to wear medical masks and disinfect all devices and workspace at the end. All this will seem obvious, but what I personally find curious is the contrast between the meticulousness of the analytical procedure, further complicated by the current Covid rules, and the simple and tangible nature of the ancient ceramics, whose immense  anthropological and material complexity, and huge archaeological interpretative potential is all enclosed in a thin section of just 30 microns thick.

In my last blog post – I introduced the method I use for the classification of the ceramic samples and the layout within the Filemaker database which I specifically designed for the purpose of the petrographic study.

So far a total of twenty-one ceramic samples from Dukki Gel has been analysed by optical microscopy (OM), while forty-three samples are currently located at the Atominstitute in Vienna where they are being analysed for instrumental Neutron Activation Analysis (iNAA) by our colleague and external expert in the project Johannes Sterba.

Most of the samples for OM (18 out of 21) are Nubian vessels among which are cooking pots (both basketry impressed and incised ware), jars, globular vessels and also fine black topped Kerma ware. Further, three Egyptian-type vessels (two red slipped bowls and one fragment of a bread mould) were analysed under the microscope.

Petrographically, the Nubian samples from Dukki Gel appear quite homogeneous in term of their composition, displaying mineralogical and textural features which also resemble very much the petrography of the Nubian samples analysed from the New Kingdom town of Sai Island (see D’Ercole and Sterba 2018; D’Ercole in prep.). Differences in the textural features, in the proportion of some specific mineral phases, and in the amount and type of the organic tempers contained in the paste allowed distinguishing four principal petrographic groups or micro fabrics. The first group is characterized by a very sandy framework with a dominant grain size in the class of silt to very fine-grained sand, a good sorting of the non-plastics and very few organics mainly small and tubular in shape. The second group also displays a sandy framework, sorting is moderate with some medium sized rounded quartz and feldspar possibly added as temper, and common tubular organics partially carbonized and moderately aligned. Group 3, to which belongs the majority of the analysed samples, is sandy, moderately sorted, with common to abundant organics, either partially or completely carbonized, heterogeneous in shape and size,  and possibly referring to various parts of plant remains (including stem, glume, palea, and lemma?) (Fig. 1) and also herbivore dung. Finally, the fourth group of Dukki Gel Nubian samples contains abundant heterogeneous organics similar to group 3 but also large carbonate inclusions of microcrystalline calcite most likely intentionally added as tempering material. To this last group, which does not show a real comparison with the material from Sai, where the presence of calcite was ubiquitous and seemed a natural component in the clay source/ soil rather than a tempering agent, refers exclusively cooking pots with basketry impressions and a single jar.

Figure 1 – Detail of organic inclusion with visible plant cell structure from Sample DG-17. PPL micropho by G. D’Ercole.

All in all, similarly to what was observed for Sai Island, the petrofacies of the Nubian ceramics is very homogeneous and points to the selection of clays, or better soils, derived from local Holocene Nile alluvia, with a composition very similar along the various sectors of the Nile river (D’Ercole and Sterba 2018). These ceramics were possibly tempered with some medium- and coarse-sized aeolian sand or with quartz grains drained by the local river systems. Technologically, the amount and type of the organic material added to the paste (more or less abundant and selected) makes the main difference and allows distinguishing among various sub-recipes or ways of doing the vessels. Further, the orientation of the voids left by the combustion of the organic matter into the paste permits to recognize among the use of different manufacturing techniques. Specifically, in the cooking pots with basketry impressions which were built on a mat, the organics appear generally well or moderately aligned with a prevalent presence of longitudinal features like stems or plant stalks (Fig. 2a). Differently, in those pots (e.g., globular pots, bowls) built with the coiling technique, the organics show mainly a poor alignment and a specific orientation that indicates the ‘relict’ coil features (Fig. 2b). The black topped and the fine polished Kerma ware generally contain less organics, these latter are also smaller in size indicating either the use of herbivore dung and/or a selection of added plant remains.

Figure 2a – Thin section scan of Sample DG-18 (Nubian cooking pot). The good alignment of the pores structures and of the voids and relicts left by the combustion of the organics indicates that this vessel was built on a mat. Image by G. D’Ercole.
Figure 2b – Thin section scan of Sample DG-17 (Nubian large bowl). The specific concentric alignment of the pores structures and of the voids and relicts left by the combustion of the organics indicates that this vessel was manufactured with the coiling technique. Image by G. D’Ercole.

Highly interesting in the sample from Dukki Gel, is the presence of a jar with a roughly polished / wet-smoothed black surface which although showing clear Nubian technological exterior features is characterized by a coarser and sandier fabric with more abundant feldspar and granitoid rock fragments resembling certain Egyptian cooking pots (Fig. 3). This sample, so far an unicum in our selection, points to an hybridization of Nubian and Egyptian traditions (this time with the intersection of some performance of ‘Egyptian’ criteria to a general Nubian technological and stylistical formula) and well supports our overall theoretical framework and working approach on the complexity and diversity among various Nubian local narrative experiences and conceptions of material culture.

Figure 3 – Sample DG-22 (Nubian jar) with roughly polished / wet-smoothed surfaces characterized by a sandy fabric rich in alkali feldspar. Photo by G. D’Ercole.

Hopefully by the end of this month, we also will have the first set of chemical data from the reference collection from Dukki Gel in our hands which we will then compare with the macroscopic evidence and with these petrographic remarks.

References

D’Ercole, G. In prep. Petrography of the pottery from the New Kingdom town of Sai, in: J. Budka, with contributions by G. D’Ercole, J. Sterba and P. Ruffieux, AcrossBorders 3: Vessels for the home away from Egypt. The pottery corpus from the New Kingdom town of Sai Island. Archaeology of Egypt, Sudan and the Levant. Vienna.

D’Ercole, G. and Sterba, J. H. 2018. From macro wares to micro fabrics and INAA compositional groups: the Pottery Corpus of the New Kingdom town on Sai Island (northern Sudan), 171–183, in: J. Budka and J. Auenmüller (eds.), From Microcosm to Macrocosm: Individual households and cities in Ancient Egypt and Nubia. Leiden.

Designing the petrographic documentation for the ceramic samples of the DiverseNile project

Documentation is the bread and butter of archaeological research. Archaeologists are daily committed to documenting everything: sites formation processes, dwellings, funerary remains, and above all the various products of material culture.

Any method of documentation, from the most essential and traditional (i.e., technical drawing of archaeological strata and finds) to the most elaborated (i.e., image-based 3D-modelling of artefacts, human remains, and sites) constitutes a fundamental step toward archaeological reconstruction. Documentation mainly serves the archaeologist to record and understand the material remains, settlement and funerary features identified during the archaeological excavation and to leave a trace of it. Also, through documentation, a preliminary process of interpretation and critical reading of the data is carried out. Furthermore, the system we adopt to document and classify archaeological data is not unbiased, rather it already implies a methodological choice and a specific scholarly interpretative approach.

As responsible, within the Work Package 3 of the DiverseNile project, for the technological and compositional analyses of the ceramic materials, I want to outline the method I use for the petrographic classification of the ceramic samples which we are going to analyse from the new concession area in the Attab to Ferka region and from our reference collections (e.g., the AcrossBorders ceramic samples from Sai Island; the New Kingdom/Kerma-Dukki Gel pottery samples; see also D’Ercole and Sterba 2018).

Generally speaking, petrography, via optical microscopy (OM), is a well-established procedure employed to examine ceramic objects and identify the source of clay raw materials and tempers used to manufacture the vessels (Fig. 1). This technique allows answering to crucial archaeological questions on pottery provenance and technology.

Figure 1. Example of ceramic thin section illustrating some common features documented for petrographic analysis. Adapted from Smith 2008: 74, Fig. 6.1.

In Sudanese archaeology, the interest in provenance and technological studies on pottery started approx. 50 years ago. In 1972, Nordström, referring to the work of Anna Shepard (1956), produced a systematic publication on early Nubian ceramics from the region of Abka-Wadi Halfa and defined the term fabric meaning the set of the compositional and anthropogenic characteristics of the ceramic material that could be determined by microscopic observation and comprised both the composition of the groundmass (or clay matrix) and non-plastic inclusions plus the potter’s technological choices adopted to make the vessel.

For the study of the ceramic material of our DiverseNile project I have designed a specific petrographic layout within the Filemaker database of the ceramic samples (Fig. 2).

Figure 2. Layout of the petrographic database designed for the DiverseNile project.

The petrographic layout includes information on the archaeological provenance and dating of the samples. It also correlates the micro fabric or petrographic group to the macroscopic evidence, that is the visual description, shape, function, and macro ware of the ceramic specimens. The consecutive entries inform on a) the groundmass or clay matrix of the sample (i.e., colour, homogeneity and optical activity); b) non-plastic inclusions (i.e., sorting, dominant grain size, maximum grain size, abundance, and mineral composition); c) plastic inclusions (i.e., clay pellets, argillaceous rock fragments etc.); d) porosity (i.e., voids abundance, type, dominant size, iso-orientation); e) organics (i.e., abundance, type, dominant size). The database also notifies on the firing regime of the ceramic sample (i.e., oxidised, reduced, reduced with narrow ox margins, dark core due to insufficient ox, oxidised to reduced). Finally, a graphic field incorporates the microscopic photos of the thin section taken under both cross-polarised (XPL) and plane polarised (PPL) light. Comments, possible comparison with other samples, and a link to the iNAA compositional groups are included as further relevant information.

The purpose of this database is to simplify the data entry of the petrographic evidence and to standardize it according to an easy-to-use, flexible, and consistent classificatory system that embraces the main information on the composition and technology of production of the ceramic data (see among others Quinn 2013).

At a subsequent step, this information will be intertwined with the results obtained from the other laboratory analyses and eventually with the archaeological data to provide a further analytical and interpretive tool for understanding the diversity and complexity of the material culture of the human groups living in the periphery of the Egyptian towns in Sudanese Nubia.

References

D’Ercole, G. and Sterba, J. H. 2018. From macro wares to micro fabrics and INAA compositional groups: the Pottery Corpus of the New Kingdom town on Sai Island (northern Sudan), 171–183, in: J. Budka and J. Auenmüller (eds.), From Microcosm to Macrocosm: Individual households and cities in Ancient Egypt and Nubia. Leiden.

Nordström, H. – Å 1972. Neolithic and A-Group sites. Uppsala, Scandinavian University.

Quinn, P. S. 2013. Ceramic Petrography: The Interpretation of Archaeological Pottery & Related Artefacts in Thin Section. Oxford, Archaeopress.

Shepard, A. O. 1956. Ceramics for the Archaeologist. Washington, DC: Carnegie Institution of Washington.

Smith, M. S. 2008. Petrography, Chapter 6, 73-107, in: J. M. Herbert, T. E. Mc Reynold (eds.), Woodland Pottery Sourcing in the Carolina Sandhills. Research Report No. 29, Research Laboratories of Archaeology, University of North Carolina at Chapel Hill.