BUTSER ANCIENT FARM ARCHIVE 1973-2007 Archivist Christine Shaw
Home History of the Project Research Publications Image Archive
Back to top

Extract from Butser Monograph:

The Discovery and Exploitation of Metals

The late A G Hamlin B.Sc.


The Butser Ancient Farm Project Trust was established in 1972 for the purpose of testing by experiment theories relating to agriculture and technology in the Iron Age (c. 300 B.C.). In the succeeding years, successful experiments on farming methods, crop yields, crop preservation and building technology have yielded data which has caused established ideas to be revised, and which has provided a sound factual base on which to assess the economy of the period.

It was realised early in the life of the Project that the pyrochemical technologies of pottery and of metal production and exploitation were important inputs to Iron Age economy which should be included in the studies. Reasonable success was achieved in the re-creation of techniques for pottery production, but all attempts to produce useful data on metal production were unsuccessful.

Butser was not alone in finding experimental reproduction of the early extraction and exploitation of metals to be an intractable problem. Many efforts have been made to reproduce the early extraction of metals by using debased modern technology in various reproductions (largely suppositional) of ancient furnaces, or by following the techniques of remote and still primitive tribes. The first approach , if it works at all, merely proves that modern technology is robust, which is not surprising as it is largely based, as far as the metals of antiquity are concerned, on technologies that are thousands of years old, but of which the chemical principles have been understood for barely two hundred years. The second approach only proves that the ancient processes from which these remnants are many stages removed were similarly robust.

So far, no technologically coherent theory explaining the discovery and exploitation of metals has been produced. At Butser, it was realised that without such a theory there could be no firm basis on which the economic importance of metals in ancient economies could be assessed. There could be no reasonable assessment as to whether metals could have been produced locally, whether they had to be traded from specialist makers, or whether, indeed, there was transition, and when, from specialist makers to local makers or vice versa. Neither could there be reasonable assessment as to when, where, and how the advances in technology revealed by metallurgical examination of archaeological artefacts could have been introduced. Butser decided, therefore, to undertake a basic re-examination of the whole problem of metal discovery and exploitation. The results are reported in this monograph.


Comparison of the areas in which Butser experiments had been successful with the area of metal discovery and exploitation in which they had not, suggested the reason for the difference.

The successful areas represented technological developments in which man had exploited the technologies at his disposal in order to achieve a desired result. For instance, given an ability to cut wood, it became possible to create a substitute for a cave at a convenient spot instead of accepting a location dictated by nature. It was therefore relatively easy for modern man to place himself in the situation of his predecessor, to perceive the desired result, and to arrive at a similar solution.

On the other hand, whoever first produced a metal could have had no idea of what he was about to produce, nor of its potential utility to him. Natural metals are rare, and even if early man had encountered native copper, silver or iron, he would have had no reason to connect them with their ores, even if, as in the case of native copper or silver, the ore had underlain the metal. Gold, which is naturally attractive, has no ore.

Discoveries, such as the production of the first metal, are technological innovations and arise when known technologies are operated in a manner that yields an unexpected result. It is therefore difficult for the present day experimenter to place himself in the position of an early man wishing to use his technology to produce a desired result when the outcome became a technological innovation. He can only speculate on what the discoverer might have been doing, given the technologies available at the time, when the unexpected occurred.

Realisation of the difference between technological developments and technological innovations explained the lack of success of archaeological experiments based on working backwards from modern technology. Man is not born with a knowledge of ore preparation, bulk charcoal production, slag technology or bellows manufacture, all of which have little or no utility before substantial production of metals is achieved. Each of these factors has to be shed at some stage in the process of working back from modern technology, and the points at which this has to be done are difficult to identify. The most likely result of such a procedure is to identify the point of degradation at which modern technology, robust as it is, ceases to work. Beyond this point is a vast area of primitive metallurgy which is inaccessible to experimentation based on modern technology.

Various avenues that might have given some insight into this inaccessible area were explored at Butser.

Many furnaces of various shapes have been excavated, and postulations regarding their original form and function have been made. Unfortunately, these postulations on original form owe much to modern technology, and one shape of furnace can perform a number of functions, perhaps even during its own lifetime.

The difficulties of studying the development of metals from archaeological remains of furnaces were felt to be further complicated by two factors which exercise important influences on the remains, but which are not easy to evaluate. They are technological secrecy and technological persistence.

It is only in the last century or so that industrial processes have been susceptible to scientific study which allows other interested parties to replicate undisclosed technologies fairly easily. Prior to this, when processes were operated empirically, there was a high premium on keeping a successful technology secret. It is therefore highly unlikely that any ancient metal working site would have been abandoned without concurrent and deliberate destruction of any features considered to be essential to the success of the technology.

Technological persistence is more insidious in that it is not what is missing from the site but what is there that may be misleading. Briefly, every change in a technology has some effect upon the installations in which that technology is carried out. These changes may not all be beneficial, some may only be thought to be beneficial, and some of the latter may actually be harmful, but once the effect of the change has been incorporated in the installation it will probably persist indefinitely and be handed on to succeeding installations. Excavated remains will therefore show features which have no relevance to the function the remains actually performed, and which, unless they can be recognised as technological persistences, will lead to erroneous interpretation of the remains. It is difficult, if not impossible, to identify technological persistences when working backwards. The matter is discussed further in Appendix 1.

It was concluded therefore that a study of excavated furnaces would be unlikely to lead to a reliable starting point for the study.

Slag deposits have received considerable archaeological attention, but it has been found difficult at Butser to see a role for slag in primitive metallurgy. Its only useful function is to facilitate the separation of unwanted mineral matter from the metal during smelting, but it can only do this at temperatures approaching the melting points of copper and iron - temperatures that could hardly be reached in primitive furnaces. If copper and iron could have been melted, then the metal could have been run away from the unwanted mineral matter in any case and slag would not have been important until the need arose to remove the unwanted matter from continuously operating blast furnaces. Nevertheless slag was produced in primitive smelting operations, and large quantities of it have been excavated. At best it must represent an unavoidable, unproductive use of fuel, at worst it may have been encouraged as being supposedly beneficial to the process, a visible sign comparable with the visible letting of blood which for centuries was a mainstay of medical practice, and which was supposed to be beneficial to the patient.

A study of slag was considered unlikely to throw much light on the origins of metal extraction for the purposes of the present study. Its most important archaeological property, melting point, which would give a guide to furnace temperature reached in its production, has rarely been measured.

Considerable effort has been devoted to the chemical analysis and metallurgical examination of early metals. The results of chemical analysis are confused by the fact that metal was man's first recyclable material, and the scrap industry must have been coeval with the extractive industry. It would seem problematical whether any virgin metal remains for analysis, nor whether it could be identified as such if it did. Metallurgical examination becomes of interest only when the extractive industry is well established, and this is substantially later than the initial discovery and exploitation of metals. There appeared to be little data in these fields of immediate value to the Butser study.