Before there was the Bowland Caves Research Project then there was the Goldsland Caves Research Project. Have you spotted the theme here? Stephen Aldhouse Green and I excavated for three years in Goldsland Wood between 2005 and 2007, working around small caves and rock-shelters and looking for evidence of prehistoric occupation. These were sites that were discovered by local amateur archaeologist Cedric Mumford in the 1980s and who contacted us and asked if we would be interested in excavating them (Ced himself was busily excavating an interesting Roman settlement at the other end of the wood). During the three seasons we were there we found fantastic evidence for Neolithic and Bronze Age burials around and within caves. This re-awakened my interest in what prehistoric people did in caves and natural places and how we might study all the different bits of the Neolithic and Bronze Age ritual landscape around these sites. Which, of course, led directly on to the setting up of the Bowland Project: which is hopefully going to give us detailed information about the use of caves, monuments and the wider environment in a way that will answer some of these questions.
Goldsland Wood itself is in South Wales. It is about 5 miles west of Cardiff on the Vale of Glamorgan, around about the place where the alien spaceship crash-landed in the pilot episode of Torchwood. The wood covers a ridge of limestone which runs roughly east-west and cutting through this ridge from north to south are two narrow valleys, Cwm George and Cwm Slatter, with small caves and rock shelters in the sides.
This photo is from 2007 and shows us excavating George Rock Shelter on the west side of Cwm George. There is no actual cave here, just a rock overhang which sheltered layers of limestone screes. Within those screes we found Neolithic pottery and stone tools, including a very nice leaf-shaped arrowhead, lots of animal bone and very fragmented human bone and teeth. The position of all of this stuff had to be recorded as it was dug out. Carl is standing by our old Nikon total station waiting to log data while, from left to right, Gemma, Sarah, Lewis, Danielle and Dave dig for treasure.
This photo shows the condition most of the bone was in when we found it. These are the partially articulated bones of a human foot. We think that people were buried at these sites by being exposed to the elements until the bodies decomposed. Lots of peoples around the world have exposure burial rites like this and we know from excavated chambered tombs that it was common during the Early Neolithic. As I posted a few weeks ago, there are lots of British Cave sites with similar burial evidence so once we started finding human bone alongside Neolithic pottery and flint tools it was always likely that the burials would be Neolithic too.
The white substance mixed up with the limestone scree is granular tufa. Tufa is a soft white rock which forms when calcium carbonate precipitates out of water in caves and streams. There was a huge layer of particularly squishy tufa around the burial deposits at George Rock Shelter. This is interesting as there are other burial caves with tufa deposits like this. Stephany Leach has suggested that in some cases in Yorkshire Neolithic people were deliberately choosing tufa caves for burial because of the petrifying qualities of the stone.
Since the dig ended we have had three radiocarbon results from human bone or teeth in George Rock Shelter. Two of them fitted very well with our expectations as they show that two of the burials took place during the Early Neolithic (between 3900 and 3700 BC in this case). The third date was on a bone from the foot in the photo above and it was somewhat more recent. Radiocarbon dating is a statistical process, so the raw result gives you a statistical likelihood of your sample having a particular date. For this reason they are usually quoted as a span of time with a probability that the sample dates to somewhere in that range. There is a 95% probability that the foot in the picture went into the rock shelter between 1680 and 1939. That’s AD 1680 to AD 1939, just to clarify.
Not only do we have good evidence for Neolithic burial practice at George Rock Shelter, as a kind of bonus we have an early modern mystery body to find out about too.
Wolf Cave was about 300 m away in Cwm Slatter. This was an actual cave but even in prehistory, before it got filled up, it was never big enough to fit lots of people inside. This photo shows the recording frenzy towards the end of the last week on site in 2007 with, from left to right, Sam, Jonny, Nicole and Rob drawing the main sections.
The finds from Wolf Cave were very similar to the kinds of things we got in George Rock Shelter and while we were on site we assumed that both caves were mostly used during the Early Neolithic. There was prehistoric pottery, lots of animal bone, including two wolf teeth – which is where the cave name came from, worked stone and human remains. This human mandible fragment came from just inside the cave entrance during the 2007 excavations. There have been two radiocarbon dates on human teeth from site A since we finished digging and they have shown that, although the artefacts were Early Neolithic, the burials are all kinds of dates. One has been dated to the Early Bronze Age, about 1500 years after the pottery we found in the cave. Another seems to be early medieval, sometime around the 7th century AD.
We are still working on interpreting and publishing Goldsland, I will be posting an outline account of the discoveries so far in the interim reports page very soon. Hopefully too, over this year we will be getting to grips in more detail with the bone evidence from both caves to try to make sense of some of the wide range of dates.