5. Roundhouses and other circular arguments

One the most fundamental and frustrating truths about archaeological excavation, tacitly accepted by practitioners, is that on most sites, a significant number of features will never be properly understood or explained.


The 1939 Little Woodbury excavation has proved to be a watershed for many aspects of archaeology of Southern England, not least because Bersu could explain and interpret the majority of the features he found, and in so doing provided the first model of the built environment of prehistoric Britain.



It is quite common for excavations to find features and datable finds without gaining clear understanding of the site's built environment. When describing what has been found, the excavator has to use an expression like ‘evidence of prehistoric activity’; meaning ‘people did things there in prehistory’. However, if buildings can be detected, we can say ‘evidence of prehistoric occupation’; meaning ‘people lived there and did things in prehistory’. Identifying buildings can raise the status of the archaeological evidence from ‘activity’ to ‘occupation’, which is much more significant and, by providing a context for other types of evidence, is vital for the overall interpretation of the site. Thus, understanding the built environment of an archaeological site is not important just for its own sake, but also because it is vital to contextualise other evidence from the excavation.


Luckily, the post-built roundhouses and other structures first defined by Bersu at Little Woodbury have proved to be fairly ubiquitous. Finding those structures that link your site into the mainstream of archaeology has become the key to a successful excavation. Having a roundhouse means your site is part of the pattern, fits in, and can be understood as part of the big picture.


However, things were never going to be that simple, because if we look only for roundhouses, that is all we will find, and we may be tempted to think that is all there is to find.


This is bad news for those sites where no obvious roundhouses can be found, giving the excavator a problem: The site won’t fit into the big picture. The problem is usually not a shortage of postholes, but rather a lack of circles, which is what fellow archaeologists expect you to find, and increasingly is the only shape they will accept for a building.

Phase plans from Gussage All Saints showing only 1 roundhouse (in phase 2) [2]

The classic illustration of this is an excavation at Gussage All Saints in Dorset, conducted by one of Britain’s top archaeologists, Geoffrey Wainwright, in 1972.[2] The site was chosen because of its similarity to Little Woodbury, but unlike the latter, it was fully excavated. The clear expectation of finding similar structures to those at Little Woodbury was completely confounded. In the end, despite the large area excavated, and the significant numbers of features found, only one small early roundhouse and 17 four-post structures were identified. Without buildings, there was not much that could be said about the nature of the site or its relationship to similar sites, and it was assumed all the buildings had been ploughed away: all very disappointing and frustrating to all those concerned.

Plans of Thorny Down Wiltshire: A. FJS Stone's Original, 1941 [3]. B. Ann Ellison's reinterpretation, 1987 [4]

The growing conviction among some researchers that Bersu’s structures at Little Woodbury could be used as a form of template led to some reinterpretation of old excavations on that basis. A site at Thorny Down in Wiltshire, excavated in the 1930s, had produced a typical plan of postholes and small pits. In his 1941 report, J F S Stone regarded localised concentrations of postholes as marking the position of ‘houses’ and other structures.[3] In 1987, Ann Ellison offered a complete, but less than convincing, reinterpretation of the site and its structures as roundhouses and 4 post structures.[4]

Two interpretations of buildings at Thorny Down: A. Chris Musson, 1970 [5] B. Stuart Piggott, 1965 [6]

Previously, in 1965, Stuart Piggot, a leading authority on British prehistory, had offered his reinterpretation of two of Stone’s houses (V & VI) as a rectangular structure.[6] Then, in 1970, Chris Musson published an alternative plan, showing how the same structure could be viewed as circular, and/or a rectangle.[5] The point Chris was making is that defining structures by ‘joining the dots’ to make shapes, without any form of corroborating evidence, is misleading and futile, a point lost on Ann Ellison.


The problem with postholes is that, if you have enough of them, you can normally make a circle with some of them. Failing that, an oval will probably do. This has led to the identification of some very dubious ‘roundhouses’, which, I am not the first to observe, is not helpful, and can result in a general loss of confidence in the credibility of archaeological reports.


However, an historical dependence on this methodology, and the uniformity of its results, has meant that roundhouses are now presumed, because no other explanation is deemed plausible. It is little wonder that archaeologists are driven to claim something even roughly circular is as a roundhouse – what else could it be?


References:


[1] G. Bersu (1940), 'Excavations at Little Woodbury, Wiltshire. Part 1, the settlement revealed by excavation'. Proceedings of the Prehistoric Society 6, pp. 30 -111.
[2] G. J. Wainwright (1979), 'Gussage All Saints: an Iron Age settlement in Dorset'. Dept of Environment Archaeological Reports No 10.
[3] J. F. S. Stone (1941), ''The Deverel-Rimbury settlement on Thorny Down, Winterbourne Gunner, S. Wiltshire'. Proc Prehist Soc 7, pp. 114-33.

[4] Ann Ellison (1987), 'The Bronze Age settlement at Thorny Down: pots, post-holes and patterning'. Proc Prehist Soc 53, pp. 385-92.
[5] C. Musson (1970), 'House-plans and prehistory'. Current Archaeology, pp. 267–275.
[6] S. Piggott (1965). Ancient Europe. (fig. 87)