There are usually more questions than answers in archaeology. Olaf has been studying the stone tools from the Goldsland Wood cave sites that Stephen Aldhouse-Green and I dug between 2005 and 2007. The worked stones from one of these sites, George Rock Shelter, give us a great example of the main question for the whole project – how did groups of prehistoric people remember significant places and ceremonies over long periods of time? At this stage we don’t have any clear answers, but we do have a whole lot of interesting extra questions, lucky us.
This is where all these worked stone tools came from, George Rock Shelter. There is a much more detailed report about this site in the interim reports section of the blog. It is a shallow rock-shelter about 6 m long where we found the remains of a least seven bodies, along with pottery, animal bone, stone tools and waste flakes.
We can look at the location of the worked stone by overlaying it onto the site plan from the interim report. The different coloured spots on this plan shows where it was all found. I have been playing with lots of different conventions on this drawing – anything with a black line around it is a piece that was burnt before it got into the shelter, the red pieces are Neolithic and the brown ones are definitely Mesolithic – I’m not sure how old the pale blue ones are. What this plan view shows best is that most of the worked stone was placed at the south end of the shelter.
However, when we do the same thing, but look at the position of the tools in cross-section, we can start to see a bit more detail. Instead of all the worked stone being part of a continuous build up of artefacts and scree, we can see that there seems to have been two distinct times when stone tools were left in the rock-shelter. Around the base of layer (1002) there are Late Mesolithic tools like this.
This is the sort of thing that was in the upper layer. The thing on the left is a leaf-shaped arrowhead. These were made in the Early Neolithic, perhaps 500 years later than the microliths from the lower layer. So far so good. Late Mesolithic flint, followed by seven Neolithic burials (we have Early Neolithic radiocarbon dates from two of the burials), followed by a final deposit of worked stone in the Neolithic. However, the piece on the right is a denticulated blade (so called because of the saw-like teeth on either side of the blade), which probably dates to the Late Neolithic, more than a 1000 years after the burials were put into the shelter.
The upper layer of stone tools is in the same part of the rock shelter as the lower one. It is made up of the same kinds of things, some waste flakes, some finished and discarded tools and some burnt pieces. How were people able to remember, over such a long time, what went into these deposits and where they should be put?
Even within these two events we can see traces of a longer history of revisiting and re-working all the different kinds of artefacts on the site. Although the blade is Late Neolithic it was made by re-using a piece from a much earlier polished flint axe.
There are two Neolithic chambered tombs very close to Goldsland Wood: Tinkinswood and this one at St Lythans. There is also Neolithic pottery and flintwork from Wolf Cave, in the next valley to George Rock Shelter. Presumably all of these sites were connected in the way they were used. And, of course, there are more and different kinds of artefact here than just the worked stone tools. One of next week’s jobs is to complete the analysis of the Neolithic pottery from the rock shelter site and see what that adds to the story.