18. Credibility Crunch Hits Iron Age Building

The story of the Iron Age building boom of the later Twentieth Century starts with one man, Gerhard Bersu [left], driven to these shores by that most clichéd of archaeological plot devices, the Nazis. It is his initial speculations and concept of a roundhouse that has dominated thinking for seventy years. The concept has become an article of faith for many archaeologists, who, perhaps unwittingly, show a near religious belief in the veracity of his ideas about buildings.

However, religion and archaeology are unsatisfactory bedfellows, and anyone who has read articles 1-9 on this site will realise that I think it is time to return to the evidence and common sense, and to prick this aged speculative bubble.
Bersu was an experienced German archaeologist, and was put to work by the Prehistoric Society in 1938 excavating the Iron Age defended enclosure at Little Woodbury in Hampshire. However, the war, and Bersu’s enforced six-year holiday on the Isle of Man with his compatriots, halted the proceedings. Despite this setback, it would not be an overstatement to say that Gerhard Bersu’s site report was one of most significant ever published for English archaeology [1].

A plan of the two superimposed roundhouses found at Little Woodbury [1]

What Bersu succeeded in doing was to take the typical profusion of postholes, pits, and gullies, an almost intractable problem for many archaeologists, and sort them out into individual structures, and to demonstrate by reconstruction the nature of the Iron Age built environment on part of the site at Little Woodbury. Essentially, he resolved a confusion of postholes into roundhouses, along with other two- and four-post structures.

It was brilliant, but perhaps due to a collective loss of confidence in the face of this ruthless German efficiency, subsequent reports that tried to reproduce his results seldom again grasped the nettle of analysing and identifying structures from scratch, from the evidence, as Bersu had done.

Two-post and four-post structures found at Little Woodbury

Therefore, since the war, archaeologists have looked for, and found, these same basic types of structures. The problem of how to understand postholes had been solved: you simply search your site for roundhouses, and so on. Your site will then be part of the mainstream, connected to Little Woodbury and all the other sites with these kinds of structures, and it gives you something to write a report about. Clearly, if everybody looks for and finds the same thing, we will end up with a nice tight data set, and the current conception that Iron Age built environment consisted of roundhouses and the other structures found by Bersu.

So what is the problem?

Firstly, all postholes that are not part of either roundhouses or four posters are effectively ignored, and in many cases this is the majority of the data. Postholes with datable material finds will make it to a phase plan, which indicates when, but not why, they were created; the rest will end up as ‘unphased’.

How the 1631 unphased features are represented in the Winnall Down Report [original plan 165 x 160mm] [2]

Since there is not much to say about what you don’t understand, archaeologists write about what they do understand; however, some reports draw attention to the data that is not understood. For example, P J Fasham’s report on Winnall Down in Hampshire notes that the 2027 postholes represented 56% of the excavated features, and that 1078 (53%) of them remained unphased [2]. Barry Cunliffe, in the report on the excavations at Hengistbury Head in Dorsett, is particularly candid about the problems of interpreting 1489 postholes, 1085 (74%) of them undated [3].

Then there are those excavations that cannot find roundhouses. This can be a disaster. Geoff Wainwright’s large excavation at Gussage All Saints Excavation found only one small one [4]. Thus the report has considerably more to say about rubbish pits than the built environment. Geoff Wainwright failed to find any buildings because he was looking for roundhouses, and the main late Iron Age building at Gussage All Saints is not a roundhouse, so he missed it.[5]

Failure to find a roundhouse affects the report's standing, as it won't fit with the rest of the evidence. Furthermore, it gives one nothing to reconstruct and illustrate; a good reconstruction illustration will greatly enhance your chances of making it into the general literature. Since, after over a half a century of excavation and report writing, we have mostly found roundhouses, (because that’s what we are looking for), it is just a small step to convincing ourselves that is all there is to find.

How complex archaeology is resolved using 'roundhouses': Springfield Lyons, Essex[6] . A: All features. B: Selected 'roundhouses'. c: Reconstruction showing roundhouses (actually a mirror image of the site plan!)
As if to complete the loop of methodological circularity, some excavators are happy to join the dots of postholes to create a variety of shapes, which are still called roundhouses. Sites like Springfield Lyons [6] and Lofts Farm [7], in Essex, are good examples, where the excavators have cracked under the pressure to conform, and reported (irregular) roundhouses! After all, what else is there to find?
The fact that we are still looking for a range of structures very similar to those found at Little Woodbury is either testament to Bersu’s extraordinary stroke of luck in finding every type of Iron Age building in his first English excavation, or else archaeologists have been content to look for what Bersu found rather than to look for evidence of buildings and structures as the great man had originally done.
Regrettably, complex patterns of postholes are difficult to understand, but archaeologists have managed to turn the study of pottery sherds into a fine art. Imagine how different it would have been if archaeology had decided early on ‘all Roman pottery was black’, and ignored all the rest. But, because ‘all Iron Age postholes form circles’, we ignore those that do not!
That many, if not most, of the postholes on some sites play no part in our understanding has to be contrasted with those forming a wall, ring beam or porch of a roundhouse, which will be absolutely central to our understanding of the site. All postholes are created equal -- but some are more equal than others.
Fig. 25 from Bersu's Little Woodbury 1940 report, the genesis of the roundhouse. [The central tower and the features on which it was based have been ignored by subsequent reconstructers.]
If a methodology dedicated to reproducing the results of a 70-year-old excavation was not bad enough, there is yet another dubious trail of conjecture and supposition leading back to the Little Woodbury report, in particular to Fig. 25, the genesis of the Iron Age roundhouse. Bersu’s Fig. 25 forms the basis for all subsequent interpretation and reconstruction, making him the grandfather of the roundhouse.
However,the undoubted father of our present-day conception was the pioneering experimental archaeologist Peter Reynolds. He used Bersu’s basic structural design for his famous experimental ‘construct’ at Butser, based on the excavation of a roundhouse at Pimperne Down.[8] Peter Reynolds was careful to call his building a ‘construct’, as initially he was trying to see if the basic model would work, as well as to learn more about the archaeology of Pimperne Down, and Iron Age building in general.

The 1976 Butser 'construct' being used as a film set.

However, perhaps because the building did not immediately fall down, the idea of ‘construct’ has been largely forgotten, and Butser became widely regarded as a ‘reconstruction’, and even a representation of an Iron Age building. Whether in the name of commerce or experimental archaeology, the many subsequent reconstructions follow this basic design. A central hearth has been added to Bersu’s model, an important and practical touch, bringing a familiar, romantic, and even emotional dimension to the model. These structures have become important as a solid representations of a vanished culture. Whether as an illustration for authors or as a backdrop for filmmakers, these ‘Iron Age’ buildings have gained significance as a device to help visualise the past.
The serious investment in Iron Age building reconstructions has signalled that the matter of structural form has been decided, or rather, was decided by Gerhard Bersu. Thus a reconstruction drawing, published in 1940, lies at the heart of visualisation of the Iron Age and, since seeing believing, a subsequent generation of archaeologists is now so fully accepting of this model that archaeological theory is being based on it. Now that everybody has seen ‘Iron Age’ buildings, it is very hard to visualise roundhouses in any other way.

But what if Fig. 25 was wrong?

There are plenty of reasons to believe that this is the case. After all, very few of the ideas developed by archaeologists in the 1930s & 40s would be accepted without question today. It is important to remember that the roundhouse born in Fig 25 belongs to a different age, where old prejudice about our ‘primitive’ ancestors was just beginning to be overturned. Proving that Iron Age people lived in substantial post-built structures, probably with thatched roofs, was actually a great step forward. Earlier in the century it was widely thought that the people had lived in small pits in the ground [9].

A 1944 Little Woodbury Reconstruction with a very primitive perspective. [10]
However, there are some serious problems with the basic model, and unfortunately the primitivism inherent in the Bersu’s roundhouse has persisted as a core belief in the study of the built environment. One particularly bizarre and unhelpful input has come from Africa, another place where circular buildings are found. Using logic similar to Erich von Däniken's linking all the world's pyramids, Reynolds and other reconstructors sought inspiration in African traditional architecture, and the resulting visual similarity is striking.

A postcard from the Iron Age or from Africa?

Why buildings built for a different environment, with different raw materials, by an entirely unrelated culture should be considered more relevant than buildings built for the same environment, with the same raw materials, and by a related culture, was never made clear. This ethnographic blind alley prevented any serious consideration of the place of Iron Age architecture in the development of thatched, oak-framed building in relation to either the previous, or the next, 2000 years. All this has contributed to the sense that Prehistory is somehow other and strange, and that that our cultural and technology ancestry owes more to the Romans than the Celts.
The failure to view timber building as a technology has divorced the Iron Age from the 5000-year-old timber building tradition in this country. The primitive notions of Iron Age building are completely at odds with the technological development and expertise shown in virtually all other aspects of material culture and, in particular, the several millennia of innovation in ‘cutting edge’ technology, the principle peaceful use of which was woodworking.
Structurally, Bersu’s design for the ring beam and roof in Fig.25 is naive, and so ‘simple’ as not to be credible as a ridged structure, certainly not one that had evolved in an environment where a lack of understanding of mechanics necessitated ‘over-design’. Simplicity in terms of roof design was still evolving 1000 years later. It eventually was achieved with the king post roof truss, still the basis for many modern timber roofs.
What primitivism tends to do is to endow previous generations with a technological helplessness, no more so than the idea that they lived in the dark, because they could not yet make holes in their buildings to let the light in. Holes to enter the building were evident, and many reconstructions even have wooden doors; however, it must have been the Romans who introduced windows to a people who had lived in the dark for over 3000 years - too backward to make holes in their buildings to let light in.
Reconstruction using a design from the 1940s, inappropriate cultural comparisons, and primitive notions of building culture has created a structure resembling a large rustic gazebo, which would have been of similar value to an Iron Age farmer, having no division of space or any obvious utility as a farm building. The only practical thing is the fireplace. However, in roundhouses like Little Woodbury, Pimperne Down, Longbridge Deverill Cow Down [11], and others, there is absolutely no evidence of central hearths!
Deconstructing the construct that gave rise to the reconstruction is not what my research is about. It is far more concerned with what is not yet understood, and the interpreting of the vast amounts of posthole evidence left over once you extract the roundhouses. This is very much a summation of several previous articles on this site (1-9), where I have sawed away at supports, and undermined the foundations of the current conceptions of the Iron Age built environment.
However, this to prepare the reader for a shock: It is not just that much of what you had thought about roundhouses is probably wrong, but also that there is an entirely different way of looking at the evidence.

In my view, having re-examined the evidence for several sites, including Little Woodbury, Pimperne Down, and Longbridge Deverill Cow Down, and having located the stairs, there is no doubt these are multi-storey farmhouses with the animals downstairs and people upstairs; there is no central fire – it would frighten the animals; and to those who might speculate that the space in these buildings is laid out according to time [12], I say ‘bullocks'!
Sources & Further Reading:
[1] Bersu, G: 1940 Excavations at Little Woodbury, Wiltshire. Part 1, the settlement revealed by excavation. Proceedings of the Prehistoric Society, 6, 30 –111
[2] Fasham, P. J. 1985: The prehistoric Settlement at Winnall Down, Winchester. Hampshire Field Club and Archaeological Society: Monograph 2. (p6, table 1 & p37, table 3).
[3] Cunliffe, B.1987: Hengistbury Head Dorset. Volume 1:The prehistoric and roman settlement,3500BC – AD500. Oxford University Committee for archaeology. Monograph No. 13
[4] Wainwright, G. J. 1979: Gussage All Saints: an Iron Age settlement in Dorset. . Dept of Environment Archaeological reports No 10.
[5] This will be discussed in a forthcoming article
[6] Buckley D G, and Hedges J D, 1987. The Bronze Age and Saxon settlements at Springfield Lyons, Essex. Essex County Council, Occasional Paper No.5
[7] Brown, N 1988: A Late Bronze Age enclosure at Lofts Farm, Essex.. Proceedings of the Prehistoric Society 54 249 – 302
[8] Harding, D W, Blake I M, and Renolds P J, 1993 An Iron Age settlement in Dorsett: Excavation and reconstruction. University of Ediburgh. Department of Archaeology Monograph series No. 1. & visit http://www.butser.org.uk/index.html
[9] References to pit dwellings can still be found in sites & monuments achieves eg:http://history.wiltshire.gov.uk/smr/getsmr.php?id=17408 & see http://www.themodernantiquarian.com/site/4150/ladle_hill.html
[10] http://www.screenonline.org.uk/film/id/469778/
[11] Hawkes, S.C. 1994. Longbridge Deverill Cow Down, Wiltshire, House 3: A Major Round House of the Early Iron Age. Oxford Journ. Archaeol. 13(1), 49-69.
[12] Pearson Parker, M 1999, Food, Sex and Death: Cosmologies in the British Iron age with Particular Reference to East Yorkshire (Cambridge Archaeological Journal 1999 (a) 9:1 pp 43-69)
(Cartoon above By B.Kliban from: The Biggest Tongue in Tunisia and other drawings. Penguin 1986, ISBN0140072209)