The full results of the phytolith work will be published over the next few years, and are already informing our excavation strategies for the coming season (if you want to know more details, you can email Tom). However, we can give some of the highlights here.
The phytolith record from Mound 1 in trench 6 at the Kaims were relatively well preserved for British samples (very little phytoliths have ever been processed from British soils, and we are the first to use these records successfully at burnt mound sites). The most striking results of our analysis is the high level of morphological and concentrational variation from different areas of the mound, horizontally across various sub-mounds, and vertically throughout these deposits. The concentrations indicate that each of the sub-mounds are deposited as distinct entities, with a consistent intensity of deposition. This implies that there is a tangible depositional sequence which starts in the south of the mound and circles round the hearth to end with the latest deposits at the north. However, the phytolith morphologies vary hugely within these distinct sub-mounds, indicating that the individual events which comprise these depositional episodes vary in character, but not intensity.
Further than these general points on depositional sequences, which actually go quite some way to furthering our understandings of how these mounds came into being and the morphologies they represent themselves in, the phytolith information can help us make some interpretations of fuel use, resource use, and burnt mound function.
The phytolith morphologies present suggest that the majority of the fuel burnt was dicotyledonous plants, and the predominance of ‘platey’ phytoliths indicates that the fuel was wood. However, in most areas, this record is accompanied by a significant level of monocotyledonous phytoliths, which represent grass phytoliths being caught up in the bark of trees prior to their firing. This suggests that the wood was not de-barked before burning. However, in some few areas the dicot level is astronomical, suggesting that the trees may have been debarked prior to firing.
The presence of varying levels of sponge spicules and siliceous diatoms within the phytolith record indicate the presence of freshwater impregnation of the deposits. However, as these two indicators are never concurrent, this points away from recurrent flooding and towards anthropogenic factors introducing water into the deposits. We know that burnt mounds are always associated with water sources, as their primary function seems to focus upon the production of hot water using heated stones. One of our current hypotheses to explain the fluctuating levels of sponge and algae within the phytolith record is that these may be being introduced through this avenue, where stones are placed in a trough of water, and then later deposited in the mound with adhering sponge and algae phytoliths.
These are all still tentative hypotheses and will be formalised with the forthcoming thin section results before publication. However, these new avenues of research indicate the importance of testing old hypotheses with new scientific techniques, and have gone quite some way to informing our investigations about resource use, function, variation, and depositional sequences within the burnt mounds in trench 6 at the Kaims.
This research is ongoing, and as such still has some distance to go before they have been fully rationalised. The next step in the process is to assess the thin section results which were taken in tandem with the phytolith samples, and then to compare the records together. But already this speedy off-season sample processing and post-ex has given us a new series of excavation strategies for our upcoming 2015 season (which you can still sign up for!). For the first time, we know that the term ‘burnt mound’ is inherently flawed! The phytolith record has shown that in fact these deposits comprise a series of individual sub-mounds as discrete entities, which in turn show enough variation in their morphologies to point towards individual events of varying character. A ‘burnt mound’, is actually a series of ‘burnt mounds’, likely deposited in episodic sequence, but with varying functions contributing to their deposition. We, at the BRP, are one step closer to cracking this enigmatic site type!