The Experimental Earthworks Revisited
Ashbee, Paul, Jewell, Peter, Antiquity
Our prehistoric monuments, the barrows, enclosures, hillforts and the like are, for the most part, earthen in that their material sources were engirdling ditches. Significant numbers are on the chalklands which have largely cradled present-day prehistory. It is only during the past century that there has been attention to the fact that such monuments are, as they appear today, a product of complex natural processes, weathering and denudation, in other words change and decay. From the outset of systematic archaeological excavation it was realized that these mounds and banks were spread, smoothed and rounded while their attendant ditches and pits, beneath their mantle of turf and mould, were infilled with chalk from their reduction and erosion. Appraisals of the timescales involved in these processes were, despite enlightened incidental observations, inexact, although it was tacitly agreed that initially they may have been speedy. Experimental controls, or contrived replication, that would give precision to interpretation was clearly a necessity. To this end two experimental earthworks were built, the first on chalk downland and the second on a gravel heath. They were planned by a committee of the British Association for the Advancement of Science that was named the 'Committee to Investigate by Experiment the Denudation and Burial of Archaeological Structures'. A detailed description of the initial endeavour was published by the British Association, entitled The experimental earthwork on Overton Down, Wiltshire, 1960 (Jewell 1963). The cover of that manual reveals the enterprise's basic theme [ILLUSTRATION FOR FIGURE 1 OMITTED]. It is in white and green, the colours of chalk downland, and it bears E.C. Curwen's depictions, published in ANTIQUITY in 1930, of the silting of old military trenches on Thundersbarrow Hill, in Sussex; superimposed in black is the precisely measured profile of the experimental earthwork. This montage epitomizes the range of the approach that the founding Committee encompassed: a backward look at prescient observation and a forward leap into scientific experimentation. This publication came to be known as the Basic Manual.
The notion of an experimental earthwork, designed for the study of silting and denudation, was proposed in a lecture delivered by one of us at the 1958 Darwin Centenary Meeting of the British Association for the Advancement of Science (Jewell 1958; 1993). Section H's recorder, John Hurst, took up the idea with enthusiasm, strongly supported by Bruce Proudfoot, one of the Secretaries, and the British Association's Committee for Archaeological Field Experiments came into being. An extended meeting of the Committee was held at the Institute of Archaeology, Gordon square, London, on 28 October 1959 and was chaired by Prof. W.F. Grimes. This meeting was memorable for its searching, interdisciplinary, discussion and, for a time, the notion of a simple structure was lost sight of for it was manifest that many barrows and earthen banks had been timber contained and laced. More intimate meetings with some six or seven colleagues, who rapidly became valued friends, led directly to an appreciation of the value of a simple bank and ditch, as well as a time-scale for its investigation. Quantification of the process to a timescale, it was felt, could bring about a new order in the excavation of prehistoric chalkland sites. We may note that in December 1957 O.G.S. Crawford had anticipated the Darwin Centenary year by publishing a special 'Evolution' number of ANTIQUITY that included Richard Atkinson's seminal paper on 'Worms and Weathering' celebrating Darwin's observations on earthworms. Evidently, thoughts of denudation and biological processes in prehistoric archaeology were current.
Besides the present writers, the Committee comprised Geoffrey Dimbleby, well known for his palaeobotany and appraisals of ancient soils, Richard Atkinson [ILLUSTRATION FOR FIGURE 2 OMITTED], notable for his work on the Oxfordshire gravels (Atkinson et al. …