Book Review

RICHARD BRADLEY. The idea of order: the circular archetype in prehistoric Europe. 242 pages, 74 illustrations. 2012. Oxford: Oxford University Press; 978-0-19-960809-6 hardback £60.

Review by Tim Ingold
Department of Anthropology, University of Aberdeen, UK

Ingold image

Why did people cease living in round houses and take up residence, instead, in rectangular ones? And why did they nevertheless enclose their clusters of rectangular dwellings within circular walls, worship in circular temples or bury their dead in circular tombs? Is there something inherently round about birth and death, from womb to cosmos, and something linear about the life paced out between them? Where circles and straight lines are co-present, what was the significance of their juxtaposition? These questions lie at the heart of Richard Bradley's masterful survey of the architectures of prehistoric Europe over a period of some six millennia, from the Neolithic to the dawn of the medieval era. Starting and ending in the middle of Ireland, Bradley travels the length and breadth of the continent, visiting sites from Spain to Turkey and from Sardinia to Scandinavia. So encyclopaedic is his knowledge of these sites and such is the authority he brings to bear in their interpretation, that the reader can be forgiven for feeling a little overwhelmed. Yet if there is one lesson to be learned from the experience, it is never to jump to conclusions. If you do, as Bradley warns time and again, they are almost certain to be wrong.

For every story of the shapes of buildings that might be read from the evidence of some sites, different stories can be read from the evidence of others. Even the same sites can be read in many different ways. Having plotted a scatter of stones or post-holes, you can join the dots to form either circles or rectangles, depending on what you want to see (p. 162). Faced with this welter of ambiguity and contradiction, anyone looking for a single master-narrative is bound to be disappointed. The lesson is indeed salutary, and Bradley is justifiably insistent that the only conclusions in which one can have any real confidence are those that are fully backed up by the archaeological record. The rest is wishful thinking. The cost of this insistence, however, is that the vast majority of potential conclusions fall by the wayside. Indeed the closing words of the book would apply equally to every step along the way: "it is not possible to tell, but it is right to ask the question" (p. 218). As the questions mount up, so our inability to tell induces a growing paralysis. All roads, it seems, lead into the long grass, and we end—as this book ends—precisely where we began, somewhere in the middle.

Nevertheless, behind all the local variations, complexities and counter-currents, there does seem to be an overall, long-term trend. In domestic architecture, the direction of transition was overwhelmingly from circular to rectangular rather than vice versa, whereas it was in the shape of monuments, from tombs to temples, that the round form was retained—all the more so as the rectangle became the norm for quotidian life. Without doubt, there is something here that cries out for explanation. Here's how Bradley puts it: "particular societies, both ancient and modern, have made a clear distinction between rectilinear and curvilinear forms" (p. 65). The point, he says, is 'deceptively simple', and yet it has scarcely been observed, let alone addressed, in archaeological writings.

I wonder, however, whether it is not Bradley himself who has been deceived by the simplicity of his formulation. Suppose that we were to extend his investigation to the monumental architecture of medieval times. Considering pillars, arches and fenestrations, we might note a transition from the Romanesque to the Gothic. The first is characterised by a combination of straight lines and circles: thus a doorframe might comprise two vertical columns capped by a semi-circular arch. But in the second, pillars resemble the trunks of trees, from which ribs fan out like growing branches to form a canopy. Wherein lies the difference? It is not that the Romanesque is rectilinear and the Gothic curvilinear, or vice versa, since in both, straight and curved lines combine in easy accord. It is rather that the Romanesque line is strictly geometric: it is defined by points—the circle by the point at its centre; the straight line as the connection between one point and another. The Gothic line, quite to the contrary, continually escapes the determination of points. It is a line of flight. This, for example, is what distinguishes the spiral from the circle, or the fan from the dome. As the Gothic shows us, 'circular' and 'curvilinear' are not the same, nor is the straightness of a column anything like that of a tree trunk. The significant difference between the Romanesque and the Gothic, then, is not between circles and rectangles but between point-determined and point-escaping lines. At issue, in short, is the relation between lines and points. What would happen if we were to reconsider the evidence from prehistory that Bradley adduces with this issue at the forefront of our minds?

Amidst this evidence is a collection of ceramic vessels from central Italy, dating from the tenth to eighth centuries BC, modelled on the forms of round houses. Not all the walls of these miniatures are decorated, but where they are, it is with "complex linear designs" (p. 47), leading Bradley to reflect, once again, upon the juxtaposition of roundness and linearity. The contrast must have had some meaning, he thinks. Yet while according some space to ceramics, he fails to mention the ancient craft of basketry, in which the zig and zag of interwoven strands constitutes the very fabric of rounded forms. Nor is this a trivial example when one recalls the claim of the nineteenth-century pioneer of architectural prehistory, Gottfried Semper, that all walls were originally woven from sticks and branches and that building, as he put it, begins with textiles. It is entirely understandable that archaeologists, excavating the foundations of buildings in the earth, should think of them as having been assembled from solid blocks rather than woven from pliant materials. But for the original builders, seeking to put a roof over their heads, it may have been just the other way around. Tectonics would have come before stereotomics, and sky before earth, in their building practice.

Indeed the contrast between the circle and the rectangle, which underlies this entire work, seems more an end-point of a process of abstraction than a starting point for one of construction. It is the abstraction of pure form that leads to the homogenisation of substance, thereby rendering pattern as surface decoration. Thus does the unity that is a woven basket come to be represented as a conical form decorated with zig-zag lines. Something rather similar happens to buildings in Bradley's account. Much of the abstraction has already happened even before the archaeologist arrives on the scene: it is effected by time itself, or more concretely by the forces of erosion, as they wear down the walls of buildings to their foundations in the ground. But it is completed by the archaeologist, in mapping these foundations onto paper. Just as the map is not the territory, so the plan is not the building. On the plan, walls become lines and posts become points. This book is full of plans, made up of points and lines. But I wonder: how can the contrast between straight lines and circles have been of such importance to the original inhabitants of the buildings, as Bradley claims it was, when it only emerges as an outcome of their ruination, and of their subsequent excavation and survey?

That this beautifully written and copiously illustrated book provokes such questions only increases my admiration for it. It is, by any account, a masterpiece. It is a privilege for the reader to join with the author in using lines and circles to think in new ways about the archaeological record. But if he could only have used the archaeological record to think about lines and circles, what further revelations might have ensued!