Today we got the results from a series of samples submitted to SUERC for C14 dating. These have provided us with some exciting new information and provided some food for thought.
Before I talk a little about the results, it is perhaps worth exploring why we decided to get some radiocarbon dates. Traditionally, Roman archaeological sites are dated using a combination of ceramic chronologies and numismatics. Combined, these can indeed provide relatively tight dating for most periods of Roman occupation in Britain. However, one of the key features about Binchester is that there is clear potential for sub-/post-Roman occupation. Activity from the 5th century is extremely difficult to date- coins are absent after the early 5th century, and whilst in some parts of the country the use of ceramics does continue to a greater or lesser extent into the 400s, there is very little evidence for significant use of pottery in the North beyond the end of Roman control and there is certainly no clearly defined chronology for it. There is also a general decrease in the quantity of other dateable material culture (personal objects; tools etc) compared with the say the 3rd or 4th centuries AD. As a consequence, unlike Roman period archaeology, early post-Roman layers are very hard to recognise on the basis of artefactual data. To make matters even more complicated in many cases, these early post-Roman layers incorporate lots of earlier, Roman material within them. In practice, this means that they often contain plenty of Roman objects, but no post-Roman objects. Not surprisingly, this means that identifying these post-Roman layers can be tricky in the extreme. One could liken it to trying to look at a lighted match held up in front of the sun – the faint traces of the flame from the former to all intents and purposes masked by the sheer power of the sunlight.
This is why we have chosen to carry out a selective programme of C14 dating on features that had the potential to begin or at least continue into the 5th century. In Trench 1, we took samples from the large waterhole (aka Hilly’s Pit) and from a stone lined pit / working hollow within the barrack block. I won’t give all the dates right now- but the waterhole produced two second /third century AD dates- we’ve already had a similar date from the same feature along with a slightly later date belonging to the 4th century. At this stage, it looks like this was a long-lived feature (given its size, this is not surprising). It probably originated in the 2nd or 3rd century AD but carried on being used into the 4th century AD. Combined with the artefactual evidence this is useful, as not only does it help refine the dating for the waterhole itself, but also to the nearby barrack. This is because the earliest phase of the stone- built barrack was partly cut by later phases of the waterhole. A better sense of the chronology of the waterhole will give us a clearer sense of when this first phase ended and the building contracted in size.
The dates from the stone-lined feature within the barrack are very exciting. There are both of mid/late 5th-6th century AD date. These are clearly and unequivocally of post-Roman date. We need to go back and look at some of the associated sequences. I do remember though that seemingly associated with this feature was a ridge of crude cobbled pavement running across the interior of the building from a wall in the north gable end. If these features are all contemporary it suggests that this northern part of the barrack at least was still standing. One job for us now is to go back and look at the associated artefactual assemblage from the pit fill. Is there anything that might be post-Roman or is it all residual? I do remember that amongst this material was a fragment of sawn antler tine. Antler working was also a feature of the ultimate/sub Roman layers identified during the excavations within the praetorium, although our C14 dates are noticeably later than these features. They could plausibly even be contemporary with the early Anglo-Saxon burials that formed the last phase of early activity in the praetorium complex. A compelling question for us now, is whether any of the animal bone dumps and spreads that lapped up around the edge of the barrack block is of the same date. Certainly, superficially, the bone-rich fill of the pit looks very similar. If the other bone is of a similarly late date, then we have a major assemblage of early medieval cattle bone with a huge research potential.
In Trench 2 our samples were all drawn from the massive dump deposits that filled in the interior of the bath-house. Here our major question is whether all this material was thrown into the building over a very short period or whether it was the result of a protracted period of dumping (over years? decades?). The dates we have back so far are all 3rd-4th century AD – this is not unexpected, but perhaps a little earlier than I was expecting. We know that amongst the ceramic assemblage from these dump layers are fragments of the distinctive Crambeck Parchment Ware (CPW). This is a very late Roman pottery type, seemingly only produced from the AD370s. One job we need to do is to go back and look at the ceramic s assemblage and work out whether the CPW is found in all the layers, including those which provided dating samples. There is still clearly more work to be done to understand this sequence and unpick the slight disparity between the C14 dates and the ceramic dates.
These are just my first thoughts about the new dates- we will try and do some more analysis in the coming weeks, and we hope to try and Bayesian modelling of the dates to refine our understanding of this new data. I’ll share our thoughts on this on the blog as usual.