The Excavator: Creator or Destroyer?
Frankel, David, Antiquity
Our special section on heritage in the June issue took the conventional current view that excavation is destruction. A more creative vision is offered. A generation ago Mortimer Wheeler articulated the basic principle that 'all excavation is destruction'. This has come to be accepted as a fundamental article of faith, and underpins the conservation philosophy expounded in the special section on Heritage in the June 1993 issue of ANTIQUITY (67 (1993): 400-445), which may be summarized in a syllogism: all excavation is destruction; destruction is wrong; therefore all excavation is wrong. I would like to respond with a contrary view that excavators do not destroy archaeological sites; they create them.
Although the impact of archaeologists is minimal when set against the multitude of direct or indirect impacts of modern society, the simple acceptance of the concept of destructive excavation can be seen in the prejudice against excavations held by many cultural heritage managers. In Australia, for example, the size of most excavations has rapidly shrunk from half the site, to 10%, to very small (50 cm x 50 cm) holes -- if excavation is allowed at all. There is little consideration given to different scales of excavation required by different research aims. Excavation is seen as exceptional, with an associated feeling that it is important to leave parts of sites untouched for future excavation, which can then 'test' the work of current researchers. While there is certainly some value in the ability to re-excavate sites, there are several problems with this argument. A policy of sampling small portions of sites rather than excavating on a larger scale has an implicit assumption of site homogeneity or uniformity; that one small part will be representative of the whole. While this may be true for the coarser scales of analysis common in much Australian hunter--gatherer archaeology, it involves a model of site formation and function which needs to be demonstrated rather than assumed. A belief in uniformity and homogeneity is also essential to the idea that future excavations can 'test' previous work by using the material from one part of a site to demonstrate that the collection of material from another part was 'wrong'. If our successors are able to excavate so much better than we can, they may never be able to relate the finer scale of their excavations to the coarser scale of ours. They will have black holes of uncertainty in the centre of their site-plans, and will curse us as much for digging small portions of sites and destroying spatial patterns as for digging the whole.
This discussion so far has presupposed that sites are identifiable and bounded entities. But, while individual as a place, a site is only one representative of a class. While some are so unusual that they must be treated as special cases, many sites conform to general patterns. There are many examples in Australia of absurd limitations where only partial excavation of one of many equivalent neighbouring sites is permitted, with the result that we have inadequate information on one site, with no data on its internal patterning of discard, and no possibility of deriving this information without digging another. It is surely better to treat the whole set of features as the sampling universe, and to dig at least one completely, or on a scale commensurate with the extent of the site.
Preserving sites for future study presumes that archaeological skill in the future will be better than the present (or how could future work test that of today?). Skills and techniques of field archaeology can only be improved by training, practice and experience; we need to excavate continually in order to assess critically the earlier field research. Knowing our skills to be limited, we must practise them in order to improve: even if this is at the expense of some of our sites. …