21 February, 2009

23.Uncovered; Prehistoric Building Regulations

I want to show you something quite remarkable about prehistoric buildings that nobody ever noticed before, so you are going to be the first to know.

The previous article, which I’m sure you’ve read, looked at the roof geometry of circular buildings, and using simple theoretical models, demonstrated that large 16-17m roundhouses are probably at the technical limits of the design. We shall extend our simple model to include a theoretical longhouse, and compare the two forms to try and understand what advantages this challenging form of roof construction offers, with interesting results.

This time we are also going to look at real examples, feeding some actual numbers into our theoretical model. As usual, truth is stranger than fiction, but two important things emerge about these ancient buildings and the people who built them, which may come as a bit of a surprise to some prehistorians.
There is an important aspect of cone geometry we have not yet covered: its centre of gravity. If you have a straight piece of circular wood, like a dowel, it should balance in the middle, but clearly, if the wood tapers, its centre of gravity is elsewhere. A bit of experimentation would establish that the centre of gravity of a cone or a tapering piece wood, like a tree trunk, is actually about 2/3 of its length as opposed to ½ if it was parallel-sided.

The positioning of the ring beam: A: Little Woodbury, correct [1]. B: West Plean, Stirlingshire, incorrect [2].
Since the centre of gravity of a conical roof, and any naturally tapering rafters, is at 2/3 of the roof diameter, this is ideal place to support the roof. It has long been known that prehistoric builders understood this, and most ring beams are positioned accordingly.[3]

Interestingly, there is an exception that proves the rule. At West Plean in Stirlingshire there was a roundhouse where the builder put the supporting posts at the middle of the roof span. This is not unreasonable in a rectangular building, but is technically wrong, both by prehistoric standards, and in terms of contemporary understanding, even in such a small building as West Plean. This serves to remind us that Mr Cockup, and his mate Mr Bodgit, are perennial characters in the narrative of building.

A simple model of the available space within a roundhouse and a longhouse of the same area.
Using a models of a roundhouse and a longhouse, we can calculate the available floor space, which we define as space with 2m headroom, including that in roof space, for a given ground floor area. The model uses a longhouse of constant width, starting with a roughly square structure with same floor area as an average roundhouse; as the roundhouse gets wider, the longhouse becomes longer. Utilising the roof space of a larger roundhouse can in theory double the available floor space in the building, adding a 2nd floor as its height increases. The equivalent fixed width longhouse has only about 50% extra space available, the roof space increasing proportionally with the length.

Plan & section of two theoretical buildings: A: a 7m wide longhouse. B: a 14m-diameter roundhouse of the same area (154m ²).
What is apparent is that the theoretical roundhouse has much greater potential floor space in its roof than a long house with the same base area. While the model is theoretical, and does not claim to show reality, it provides some structural rational for the construction of such a wide roofs.

A Case Study
Our case study involves taking some real building dimensions and feeding them into our models. This is where we encounter a bit of a problem with evidence, in that walls rarely survive. Prehistoric roundhouse walls were not usually load-bearing, and had shallow foundations, so they are often ploughed away.

Case study: Theoretical models for real archaeological buildings

It is worth noting that the diameter of the roof should include the ‘eaves’ (the part of the roof that overhangs the wall), which is the widest and heaviest part of the roof. However, this difficult to detect archaeologically, so for our case study a nominal 1m is added to the wall diameter to give a figure for the roof base.
Roundhouse plans: A: P10 Moel y Gaer; Clwyd[4] B: Pimperne Down, Dorsett [3] C: Little Woodbury, Wiltshire[1] D: House III, Longbridge Deverill Cow Down, Wiltshire [5]
This problem of archaeologically invisible walls was highlighted by Graeme Guilbert, whose excavation at Moel y Gaer, a Welsh hillfort in Clwyd, provided an large assemblage of roundhouse plans, including building P10, where evidence for a stake-built wall survived.[4] The other examples have evidence for the position of the wall, and have all at some time have been reconstructed, and will be familiar to readers of my previous articles.

It is probably not a coincidence that that the largest two buildings have post-built, as opposed to stake-built, walls. This is probably because some of the roof mass will inevitably act on the wall, and a post-built wall has greater load-bearing potential. A rafter resting on circular wattle and daub will effectively only bear on that part of the wall it is in contact with, it is a ‘point load’, whereas a post and lintel wall can distribute any load between its posts.

Case study: Theoretical model roofs for real archaeological buildings
P10 is one of the larger structures found at Moel y Gaer, and visually not that different from the other examples, but compared with the Cow Down, its roof is half the size and weight. Its rafters, at 8m-9.5m long, would not be as difficult to grow as the 11.5m –13.5m timbers required for the largest example.

Case study: Theoretical model of volume and floors area for real archaeological buildings

A modern three bedroom house might have a total floor area of about 150m ² /1600 sqft , so Moel y Gaer is smaller than this, but Little Woodbury is larger on the ground floor alone; including the upper floor, it could be twice this size. The difference a small change in diameter can make to a roundhouse is emphasised again by the figures for theoretical building volume and floor space, Cow Down being twice the size of Pimperne Down, and 3 times the size of Moel y Gaer.
Case study: Theoretical model of post loading for real archaeological buildings
The big advantage with the real examples is that we know the number of posts used to support their theoretical roofs, so by simply dividing the mass of the roof by the number posts, we can calculate the loading on each post, both for the dry weight and for a snow-covered roof.

What is so remarkable is that the figures for post loading are almost identical for three of the buildings, and just slightly lower for Pimperne Down, which indicates that the builders were using similar principles of foundation design. We have seen the significant difference, in terms of roof weight, between the largest and smallest buildings, so to find that the actual loading on each post is so similar is good news, but actually what you would expect if you were dealing with specialist architects.

The figures are an estimate derived from the model, and the parameters can be varied or refined to produce different figures. However, what is important is that the same model is used on each. In general terms, they are quite conservative figures, quite comparable with the modern figures we considered in a previous article,[6] and similar to figures for other prehistoric buildings I have studied.

Understanding prehistoric foundation design is particularly useful when we are considering the underlying problem of how to identify buildings from the evidence of postholes. Knowing you are looking for the work of architects who understand the importance of level foundations and utilise a rational system of design is a great help when you are trying to detect structures. Theoretical structural archaeology is a methodology, a way of looking at the evidence, which sees postholes as foundations, and tries to detect and understand the structures they supported, with the broader objective of understanding the development of prehistoric architecture in these terms..

A photograph of the excavation of the second phase building at Pimperne down.
The Pimperne Down roundhouse is most conservative, with its 27 posts having the lowest post loading. However, when this structure was rebuilt, the second phase has fewer posts (21 or 22), increasing the dry loading to close to that of the other buildings.

Rebuilt roundhouse plans: A: Pimperne Down, Dorsett [3] B: Little Woodbury, Wiltshire [1]
Little Woodbury was also rebuilt and its 25 post ring beam replaced by 21 posts. This suggests that prehistoric builders had safe working limits, which they continued to refine, trying to create the maximum amount of space on the minimum of safe foundation, the same process which is evident in architecture throughout history.

The big story here is that prehistoric timber building was architecture created by competent specialists. Thousands of years before the appearance of building regulations and the man from the council, prehistoric builders had it covered.

Sources & Further Reading:

[1] G. Bersu, 1940. Excavations at Little Woodbury, Wiltshire. Part 1, the settlement revealed by excavation. Proceedings of the Prehistoric Society, 6, 30 -111.

[2] K A Steer, 1956. "An Early Iron Age Homestead at West Plean", Proc Soc Antiq Scot vol.89, pp.227-49
[3] D W Harding, I M Blake, and P J Reynolds, 1993. An Iron Age settlement in Dorsett: Excavation and reconstruction. University of Edinburgh. Department of Archaeology Monograph series No. 1. Also: http://www.butser.org.uk/iaflphd_hcc.html & http://www.butser.org.uk/index_sub.html
[4] Graeme Guilbert, 1981. Double-ring roundhouses, probable and possible, in prehistoric Britain Proc Prehist Soc 47, 299

[5] S. C. Hawkes. 1994. Longbridge Deverill Cow Down, Wiltshire, House 3: A Major Round House of the Early Iron Age. Oxford Journ. Archaeol. 13(1), 49-69
[6] http://structuralarchaeology.blogspot.com/2009/01/20-everything-you-ever-wanted-to-know.html


tim said...

Hi Geoff

Still yet another excellent post love the photo at Pimperne down.

Geoff Carter said...

Thanks Tim,

If you want good site photos - dig on chalk, it looks great in black & white!

alice said...

Hi Geoff, fascinating articles over the last month - I've really enjoyed thinking along with your observations in terms of my own roundhouses in the Caribbean. I wonder have you ever taken a look at some of the vernacular architecture in the Amazonian area? this is basically the source of inspiration for reconstructions of the very few houseplans we have from the islands -south american malocus. these people (yanomami, yekuana, achuar and hundreds of other societies) build amazingly elegant and complex post and palm structures in a huge range of sizes and styles - some durable and symbolically laden, other ephemeral and profane shelters. the great thing about the ethnographies is that they sometimes describe in detail the building sequence, physical and resource constraints, and indigenous belief systems behind the architecture. anyway, I'm going to get back to thinking about roof pitch...by the way, if you have a beehive-shaped house, i.e. not a cone on a cylinder, but a domed frustum (good word!) with the wall posts in the ground and joined at the top to also form the roof, do you have any idea how this changes the weight distribution throughout the structure? does it all get thrust into the ground?

Geoff Carter said...

Hi Alice, Thanks for the comment.

In theory, that is the advantage of domed structures, but the performance of a domed structure depends on how it is constructed, the materials, and, as always with circular off shapes, its size. The ability of the structure to restrain the trust will also depend on its rigidity; small domed structures like wikiups, made of flexible branches or split wood are strong enough on a small scale, but they would not necessarily ‘scale up’. The competing requirements of flexibility, to enable construction, and rigidity, so that the dome does not deform under load, were never satisfactorily resolved using wood, and was finally resolved by the Roman use of concrete.

It is important to consider each building individually, and not to generalise on the basis of shape, - especially with circular buildings. Archaeologists are always the first to complain when people make connections between the world’s pyramids, for example, simply on the basis that are similar in shape, humans and archaeologists seem drawn to the simple explanation, and the one-size-fits-all explanation, (previously offered by religion).

I am always interested in, and happy to look, at building plans, and if their cultural context is understood that saves a lot of speculation.

The next post is about how to distinguish sacred and profane space in Celtic structures, so stay tuned!

AJ said...

Nice work Geoff, and so much still to explore..

Geoff Carter said...

My pleasure AJ, thank you so much for taking the time to read it.

NB: Your site is well worth a visit from anyone interested in Stonehenge.

Anonymous said...

Pre history is so interesting; the further back you go, the more we progress to the future..

Why did the Inca use stone? And basalt at that. My own theory is that the original inhabitants of Lake Titicaca were the survivors of an environmental cataclysm that propelled their move to the highest available area from their destroyed homeland, the site of the Cayman Trench. Note that the mysterious rip in the ionosphere looms above. Much later some moved to a more pleasant clime, Crete, selected because it wasn't volcanic!!

These groups seems to be connected by their preference to the 'double axe', which I interpret as a double axis, as the parallel faults running through the island of Hispanola. Note the Inca recreated this in their roads, one running along the mountain ridge and one along the coast. The Minoans were known as the land of the double ax, even though an axe played no part in their culture.

So, I suspect their relationship is linked to the destroyed civilization in the Cayman Trench area.

I have Knols on line at google if interested.
Thanks for your site.

Still, why cut basalt?

Geoff Carter said...

Much of the prehistoric world was made of wood, and it was the also principal fuel for heating, cooking, and manufacturing. Axes are one of the first and most important tools in the human tool kit. For cutting, splitting, and fashioning wood, you require a hard axe, with a durable edge.

Basalt is used for axes and other stone tools because it is very hard, and importantly, it is fine grained and so can be ground into the required shape, using another stone and abrasive sand. It is hard work, but ideal for slaves.

The Inca civilisation was flourishing over two thousand years after the Minoans, there is really no need to link civilisations in this way; there has always been clever and inventive people who have original ideas.

The wheel was also ‘invented’ in Mesoamerica, but possibly because of a lack of obvious draught animals, it was never developed a means of transport.

Kaloy said...

The way the ancient world built their building is somewhat a great mystery to many, well, for the fact that they have less resources way back then, but funny how their building regulations are somehow on a professional basis, I mean how could they be able to build awesome architectural designs if building regulations does not exist during their time, right? haha. But anyways, time has changed the ways man live, but there are still some that remains, the idea of following building regulations for example, has lived throughout the century, although of course some changes for the better has been made. Thanks for sharing.

Geoff Carter said...

Thanks for your comment Kaloy;
Craft is a wonderful thing, mostly based on the experience of previous generations.
However, we only see the ancient buildings that were well built, the medieval cathedrals that fell down are now forgotten, along with their architects!

Renny Barrios said...

Excelent! Reallly interesting blog

Geoff Carter said...

Thanks, - I think prehistoric buildings are very interesting !

Becky Smith said...

Hi there, Brilliant, brilliant article! Is it published anywhere? Id like to reference it if I may?

Geoff Carter said...

Thank you – very kind of you to comment, I am so pleased when somebody gets it; more than happy for you to reference it.
All this work has a stable Google URL; it should be here for ever!
One day I would love to ‘publish’ this data, it is part of a larger study I undertook; I am trying to write a book.
[It was to be part of PhD – but it is evidenced based, and a not compatible with Post-processual archaeology].