A new methodology

July 15th, 2017

General introduction
This article describes the author's methodology for the analysis of Attic Geometric and Orientalising finewares, which was originally included in his doctoral thesis in 2015. The article below is a revised version of his original methodology, adapted for the needs of an online article. The author has also decided to include his critique on some aspects of his own work, which was not originally included in his doctoral thesis.

Which aspects of the chaîne opératoire theory were examined by the author and how?
According to the discussion on the chaîne opératoire theory, a main concern in the analysis of archaeological ceramics is the aspect of technological choice. Although technological choice manifests in different domains of ceramic production and is subject to a range of social parameters, the author's research focused on three of its most practical aspects. These were the conceptualisation of ceramic forms, including topology, partonomy and sequence (sensu Van der Leeuw 1994, 136-7), raw materials, and finally decorative technology. The above aspects related to the four basic attributes (fabrics, shapes, dimensions and decoration), which according to Kotsonas (2014, 1) define standardisation or variation of ceramic materials.

Through the analysis of the above aspects, the author's research project investigated artefact variability (sensu Schiffer & Skibo 1997) across three broader groups of Attic Early Iron Age finewares: large ceramic containers, medium sized pouring vessels and small drinking vessels. Such analysis of artefact variability was likely to suggest two possibilities: Firstly, if a specific ware showed no variability across time, then its production was probably regulated by strong technological traditions (sensu Sillar & Tite 2000), leading to standardised forms. Standardisation in the production of a specific shape may have indicated specialisation within the production sequence. Secondly, if a specific ware exhibited some degree of variability across time, then its production could have indicated absence of distinct technological traditions in favour of experimentations, innovations, or freedom of technological choice among potters, leading to non-standardised ceramic forms.

As explained in the articles on the chaîne opératoire approach, artefact variability is related to cognitive and perceptual motor competences (sensu Roux 1990, 142) related to the potter’s behaviour. According to the Behavioural Chain Analysis Theory (sensu Schiffer 1995, 57) a potter’s behaviour can be examined by reversing the sequence of production of the ceramic artefact; therefore, the final focus of the author's original study was the potter’s behaviour and not the artefact itself. The rationale of the author's work is summarised in the figure below:

Conceptualisation (sensu Van der Leeuw 1994, 136-7) was studied through the analysis of metrical features of ceramic vessels and the mathematical proportions between them. Particular interest was placed on vessel height. The process of recording metrical features of ceramic vessels was straight forward: measure tapes, rulers, callipers and diameter charts were used to obtain characteristic measurements of a vessel’s shape. In relation to partonomy and sequence (sensu Van der Leeuw 1994, 136-7), complex forms such as Early Iron Age amphorae used to be assembled from a number of different constituent vessel parts. Unfortunately, the identification of such constituent parts was not a straight forward process. Intact wheel-made finewares do not reveal their constituent parts easily because joints between them do not survive. If the assembling procedure has been executed on the fast wheel, some surface marks may exist; however, one cannot be fully sure. X-ray analysis often reveals such joints and points out the existence and exact number of assembled parts. For example, Josef Noble (1966, 24, 155, fig.150) used X-Ray radiography to show that Athenian Classical lekythoi where produced from a single piece of concrete clay, which was then drilled to formulate the inner cavity of the vessel.

In the author's original project, the selection of pottery that allowed visual identification of constituent parts needed to be planed carefully because of various restrictions. Firstly, intact vessels appropriate for such analysis were difficult to find. Areas with long term occupation such as the Athenian Agora had primarily produced fragmented pottery. Sites such as the Kerameikos cemetery had produced a large number of intact vessels; however, the most important ones were at the time in museum display or stored in facilities with limited access. A primary concern during the project was that vessels located in museum collections could not be easily removed from display. Secondly, X-ray analysis required a time-consuming process of acquiring scientific analysis permits from the local authorities. Vessels required to be transported to research facilities outside the museums; therefore, special arrangements were necessary for their safe transportation. Due to lack of resources to facilitate such procedures, the study of constituent vessel parts was carried out through visual examination of fragmented vessels, which revealed visible joints on their surfaces. Access to such material was also quicker and safer for the artefacts. Macroscopic analysis showed that the only two constituent parts safely identifiable were necks and handles; therefore, the study needed to be restricted in the metrical features and proportions related to those two vessel parts only.

Despite this strategy, a second problem arose: smaller vessels such as oinochoai and skyphoi survived in better condition in the archaeological record compared to larger ceramic containers; therefore, sample numbers would have favoured an analysis that focused on medium to small-sized pottery. To overcome the problem and to improve statistical accuracy, the material studied by the author needed to expand. It was decided to include two additional types of pots: firstly, vessels restored up to a good degree preserving key metrical features; secondly, mended or partly mended vessels preserving complete profile. The study of complete vessels that were at the time in display was conducted though published photographs. To limit any bias, a 10% sample of these vessels was selected and studied macroscopically. The results of the macroscopic study were tested against the results of the study conducted through published photographs and showed insignificant and limited bias.

Despite the major focus in investigating artefact variability in relation to potters’ technological choices, the author's discussion could have not been limited to the work of one group of artisans. Equal attention needed to be placed on the work of painters, particularly because the majority of studies on Attic Early Iron Age workshops are based on iconography and connoisseurship. During the original project, pottery decoration was analysed as a technological rather than stylistic choice: decorative variability was inferred from the repetition of trends in the use of specific external treatments, instead of the repetition of specific decorative motifs. More specifically, macroscopic analysis of painted colours, slips and coatings was used to define patterns of continuity or interruption in decorative practices across time. Paints, slips and coatings were not just seen as decorative features, but also as technological features: their external appearance would have varied according to their chemical composition and the effects of firing.

The final concept in the investigation of technological choice was natural resources. In ceramic production natural resources are connected to core manufacturing processes such as clay selection, clay manipulation, tempering and firing (Rise 1987; Sinopoli 1991). Scholars of the chaîne opératoire approach usually discuss natural resources through ethnographic research (e.g. David & Krammer 2001); however, in a study of archaeological ceramics, macroscopic (hand specimen) and microscopic (archaeometric) techniques are the most popular in the investigation of fabrics.

The most convenient fabric analysis technique is commonly known as hand specimen examination. It is widely used by field archaeologists who require a fast and pragmatic fabric description to supplement their work. Hand specimen examination is performed by visual analysis on a sherd’s fracture with the use of a low magnification (X10) hand lens. This examination offers colour descriptions based on the Munsell Soil Chart, identifications of voids and tempered inclusions in relation to their frequency, size, sorting and rounding, and finally information regarding texture, feel and hardness of a sherd’s fracture (Orton et al. 1993, 231-41). According to the Greek antiquities legislation, fresh fractures could not be produced on ceramic artefacts without applying for a destructive analysis permit; therefore, the author's study of fabrics had to be restricted in artefacts that were already fragmented and allowed instant hand specimen examination.

The author's work also included an archaeometric pilot study on Athenian Geometric and Orientalising sherds. The study was planned to analyse fabrics with the use of archaeometric techniques such as Thin Section Analysis (TSA) and Scanning Electron Microscopy (SEM). The pilot project was carried out independently and parallel to the macroscopic analysis of larger assemblages discussed in the author's thesis. This small assemblage of 17 unpublished Athenian finewares came from the same contexts as the rest of the published material analysed macroscopically, but not from the same vessels. Due to restrictions in Greek antiquities’ legislation, sampling of published artefacts was avoided.

A similar approach with the use of combined hand specimen examination, ceramic petrography and Scanning Electron Microscopy was carried out by Hilditch (2014, 32, fig.3) for the analysis of the ceramic chaîne opératoire in the production of Minoan conical cups. Despite some problems of clarity in relation to her sample sizes and the general concept of ferrous clays, Hilditch (2014) demonstrated that microscopic techniques are highly useful in mapping ceramic chaînes opératoires.