The influence of post-burial conditions on trace element composition of ancient sherds

Abstract

In provenance studies it is normally assumed that the concentrations of elements used for diagnostic discrimination are not affected by post-burial conditions. We analyzed 28 well preserved sherds, stored under wet conditions in a Roman garbage pit at Cologne. We also analyzed 79 sherds from assorted garbage pits and other sites about Cologne. These sherds included both well preserved and poorly preserved specimens of different types of wares. All of them were made available to us by Dr. R.C.A. Rottlander (Tubingen). As well, artificial leaching experiments were conducted. The results were augmented by optical and mineralogical analyses. The study showed that the variations in elemental concentrations among and within the subgroups did not correspond to changes in the preservation of the objects.

Post‐Depositional Elemental Alterations in Pottery: Neutron Activation Analyses of Surface and Core Samples*

“…Examples of this approach are provided by, for example, Picon (1976, 1991), Franklin and Hancock (1980), Lemoine et al (1981), Rottländer (1981, 1989), Buxeda (1999) and Buxeda et al (2001). The dependency on the archaeological hypothesis, however, makes this approach risky—even if the assumption of a common origin is true, the potter might have changed his recipe for the production of the clay paste in the course of time, which would result in different ‘chemical fingerprints’ within the same workshop.…”

On the Source of Clay Used for Cologne Roman Pottery

“…The exceptions are the concentrations of elements such as Sc, V, and Ga, which lie close to a 20 elevation in the Frechen potter's sherd, and Hf which is similarly elevated in the Frechen spacers. These differences are not sufficient to differentiate between the two very small groups, but indicate that the 'modern' Frechen wares might exhibit the same type of interelement variation as was observed in the Roman wares (Franklin and Hancock, 1980). Bearded man jug sherds were excavated from Frechen and Cologne but were probably manufactured in Cologne using Frechen clay (Rottlander 1979).…”

“…very low calcium wares. The anomalous behavior of Hf, which increases in concentration from Roman to 'modern' wares, tends to indicate tempering by zircon-accumulating sand temper, since the Hf has been shown to correlate positively with Si in Roman wares (Hancock 1982) and negatively with most of the other major and trace elements (Franklin and Hancock 1980). Therefore, given the proposition that the 'modern' wares have extra sand temper relative to the Roman wares, one would expect interelement scattergrams to exhibit interelement correlations which are:…”

“…Some aspects of the chemical composition of the pottery produced in Cologne during Roman times (Rottlander 1981) have been studied using neutron activation analysis of sherds which were part of a large collection (Franklin and Hancock 1980, Hancock 1982, Rottlander 1982a, 1981b. Although these Roman pots were known to be manufactured in Cologne, the source of clay used in them has not been firmly established.…”

Thoughts on the Utility of a Bivariate‐splitting Approach to Olmec Ceramic Data Interpretation*

“…On occasion, there have been debates as to which data interpretation is more realistic (e.g., Franklin and Hancock 1980; Rottländer 1989) and may lead to more convincing proposals concerning the movement of ceramics around specific geographical regions. One recent such debate centred on the interpretation of a major set of INAA data based on Olmec ceramics from Mexico, generated at the Missouri University Research Reactor (MURR), and published recently.…”