TL-dating of vitrified material

Abstract

While attempting to date material from the vitrified ramparts of various hill-forts in central and western Europe by thermoluminescence (TL) methods, obviously erroneous results appeared. Searching for possible reasons, an investigation into the upper temperature limit of TL-dating was started. It is well known that insufficient heating causes far too old erroneous TL-ages. Less well known and yet unexplored is the temperature region where vitrification commences. From a hill-fort near Uppsala, samples were taken along a trench through the vitrified rampart. Firing temperatures were determined for each sample by petrographic methods within ± 5 0 ° C, and TL-measurements were carried out. The results show that the ages determined become successively younger with increasing firing temperatures, for samples heated to temperatures exceeding 900°C. The younger the apparent age, the higher the glass content of the sample. Although most of our data are from vitrified ramparts - somewhat exotic to the majority of archaeologists - our findings apply to all strongly heated materials. Examples are archaeometallurgic remains (slags, furnaces, hearths), tempered ceramics (stoneware, porcelain), or molten bricks. The paper also presents clearly erroneous T L - dates for several of these materials, along with an explanation of the phenomenon.