Debunking the myth of timber circles

The Myth

In archaeology there is the evidence, and there are the opinions put forward to explain and connect it.

In the highly subjective world of archaeological opinion, it has become widely accepted that the concentric rings of prehistoric postholes, once thought possibly to be buildings, were the freestanding timber equivalents of Neolithic stone circles.

These ‘timber circles’ have become central to modern understanding of ancient landscapes, as realised by Time Team. [1] Recently, Professor Vince Gaffney has even detected a ‘Timber Stonehenge’ [2].

They account for some important archaeological sites including Woodhenge, Durrington Walls, and The Sanctuary, and are even found in the Iron Age.

In 1996, Tim Darvill called these structures Type Ei buildings, [3][above left]. However, once others, particularly Alex Gibson, ventured the opinion they were too large to be roofed, they became ‘Timber Circles’ [4]. Many academics have long since moved off to explore this strange freestanding ‘ritual’ landscape, although, officially, it appears that the jury is still out:

“A timber circle is the foundation of a large wooden structure comprising a series of two or more roughly concentric rings of postholes which once supported substantial timber uprights variously interpreted as stanchions of a roofed building or freestanding posts. Timber circles are generally over 20m in diameter and the individual postholes are typically over 0.5m across.”

English Heritage web site. [5]

Actually, they are not too big, but entirely consistent with the scale other large oak timber roofs, and this can be demonstrated because they were made from trees.

Please click to enlarge the drawings, and consider them; they form an important part of an argument, which centers on how to understand archaeological site plans.

The Reality

In comparison to medieval roofs, they are not too wide, and have considerably more supporting posts; most are 50' /15.24m wide or less. Westminster Hall, a timber roof on stone supporting walls, with no posts, is 68’(20.72m) wide. [Above].

In essence, most type Ei buildings took the form of a very wide Neolithic longhouse with a central supported ridge and a symmetrical arcade and roof plates. They usually comprise five concentric rings of posts, forming an annular shape. These posts are arranged to support a polygonal ridged roof that narrows on the inside and widens on the outside.

How the center worked is more difficult. Some, like Woodhenge, with a centre narrow enough to be roofed, have an extra sixth set of posts inside the middle ring.

The middle ring, making the ridge of the roof, has the largest, deepest postholes, as these posts would be the tallest vertical components of the building. The postholes become progressively shallower away from the ridge, marking the shorter posts supporting the arcade and roof plates.

Traditional timber buildings were made from the trunks of oak trees, the growth pattern of which is broadly predicable [right]. In general terms, the longest usable tapering timber that these trees will produce, neither too fat at one end, nor too thin at the other, straight and without branches, is about 50’/15.24m. The availability and natural limits of these trees can be optimised through managed woodland, but ultimately they dictate the scale of individual timber components, such as rafters and ties.

A tie is a piece of timber running across the base of a roof constraining the outward forces in a large roof. It is subject to the same limitations as other timbers, and this in turn, limits the width of the roof. This explains why these structures, like other large timber roofs, including Iron Age Roundhouses, are limited to about 50’ wide.

This is not the only factor affecting scale in timber buildings, and like all architectural limits, it must expected that it will be pushed. At the same time, since a 50’ timber is probably comparatively rare, that available for ‘ordinary buildings’ will be significantly smaller, in the 30–40’ range.

The posts supporting a 50’ tapering tie should be large at one end and small at the other; this explains why the postholes of the outer rings are larger than the inner ones [above]. If all the horizontal timbers are placed with their thin end pointing inwards, this will simplify the assembly of the frame, avoiding having to join components of radically different thickness. This is also reflected in the generally smaller postholes on the inside of the middle ring.

A 50’ tie would require support at least at one point, and so, if you draw in the line of a tie between corresponding posts [holes] in the outer and inner ring of a class Ei building, you will find it will usually pass next to one, or more, internal posts. Looked at in more detail, because there are a greater number of outer posts than inner ones, a pattern of radial ties, laid out much like a bicycle wheel, is evident. [Left]

Interlace theory takes these ideas much further, beyond the effect of the tapering timber on the plan, into the detailed structure and geometry of the roof.

These are some of the largest timber post structures in prehistory, so it is entirely appropriate these buildings are constructed at the limits of design, and not beyond it. Each example is individual, and clearly the product of architectural and building techniques used to furnish large roofed structures where society and individuals required it.

Understanding in conflict with belief

‘Timber circles’ actually are buildings, as Professor Darvill suggested fifteen years ago, the posthole foundations of top-of-the-range oak timber architecture. With building, as with most crafts, some clients demand the best available, which in architecture is often the biggest, tallest, or widest. Admittedly, they are complicated and not straight-forward, but that is precisely because they are big buildings made from hundreds of tapering tree trunks.

This explanation requires no belief, only understanding. It accounts directly for all the basic characteristics of the evidence, explaining the size, depth, and distribution of the postholes. It does so in terms of the typical form of the trees evident in the construction, and their position in the geometry of the structure.

Trying to connect these complex timber structures with circles of stones, largely on the basis that they are both circular, does not really wash as grown up argument. It has led to much barking up the wrong trees, and the myth of timber circles, a tale that has grown much in the telling, so much so that those who say “archaeologists make up stories about the past” may have cause to feel vindicated.[6]

Next Related Article: Stonehenge and the archaeology of the prehistoric roof

Sources and further reading

[1] http://www.channel4.com/history/microsites/T/timeteam/2005_durr.html [accessed 6/03/2011]
[2] http://www.telegraph.co.uk/science/science-news/7903824/Ancient-wooden-version-of-Stonehenge-found-on-Salisbury-Plain.html [accessed 6/03/2011]
[3] ’Neolithic houses in northwest Europe and beyond’’ (1996) Neolithic Houses in North-West Europe and beyond (Oxbow monograph 57) [Paperback]. T.C. Darvill (Editor), Julian Thomas (Editor) [Fig 6.9]
[4] Gibson, A. 2005. Stonehenge and timber circles (revised 3rd edition). Stroud: Tempus Alex M. Gibson
[5] http://www.eng-h.gov.uk/mpp/mcd/sub/timc1.htm [accessed 6/03/2011]
[6] http://www.timeshighereducation.co.uk/story.asp?storycode=404899 [accessed 6/03/2011]