In my previous incarnation, I was Education Interpreter for the Celtic Village here at St Fagans. One of the last projects we undertook before I left was to consider the possibility of re-building one of the houses with a turf wall rather than with wattle and daub.

Below, you will find a brief outline of the reasons why we think this is archaeologically sound. What do you think? We will try to engage as many archaeologists as possible in this debate by sending out the link to various contacts. We then hope to organise a mini conference in 2008 to discuss the pros and cons of turf walls. Please feel free to contribute, and please send the link on to more people.

Understanding Welsh hillforts: Experimental reconstruction of a turf-walled roundhouse

The Celtic Village at St Fagans: National History Museum, Wales is a popular attraction for visitors, schoolchildren, students and archaeologists. It comprises two wattle and daub structures, based on evidence from Moel y Gerddi, Gwynedd and Moel y Gaer, Flintshire, and one stone structure based on evidence from Conderton, Worcestershire. They were built in 1992 by Dr Peter Reynolds, the leading expert in roundhouses at the time, and have been preserved, with essential modifications, ever since.

However, the Moel y Gaer reconstruction is nearing the end of its life, just as predicted by Dr Reynolds. We therefore propose to demolish the present structure, and to rebuild it, basing our efforts on current archaeological theory. We are interested, therefore, in starting a discussion with leading experts in the field, with a view to establishing a networked web forum of interested parties which will share ideas and advice on general roundhouse reconstruction.

We are presently in discussion with Dr Rachel Pope, Lecturer in Prehistoric Archaeology at Liverpool University, about her research into Iron Age roundhouses. With her advice, we are interested in re-building the Moel y Gaer as a turf-walled structure.

Dr Pope’s doctoral research – funded by the AHRC – combined a large-scale dataset of 1180 prehistoric houses, with a long-term trajectory (2400 BC-AD 500). All aspects of prehistoric roundhouses were studied - their design, construction, use and lifespan.

She has also studied the Welsh Roundhouse – a project funded by the Board of Celtic Studies. From this extensive research, she proposes that stake walls were used as a dominant form in reconstruction largely due to the misinterpretation of structural remains from Wessex e.g. the Longbridge Deverill and Pimperne reconstructions at Butser, both of which have failed due to structural problems. The Moel y Gaer structure has also failed, with the wall bowing at the entrance at a rate of approximately 1 cm per year.

Dr Pope argues that convincing evidence has emerged which suggests that the structures were wide turf-walled structures with a wattle lining, rather than having a stake wall and projecting porch. Turf construction is well understood in Iceland, with the reconstruction of Norse longhouses and traditional turf dwellings of the eighteenth century. It is also found in northern Scotland with some traditional blackhouse.

The research is supported by Richard Brewer (Keeper of Archaeology, National Museum Wales), Gerallt Nash (Senior Curator, National History Museum), Adam Gwilt (Later Prehistorian, NMW), Owain Rhys (Curator of Contemporary Life, NHM), Ken Brassil (Learning NMW) and Nia Williams (Learning, NHM).