Refuse and the Formation of Middens
Needham, Stuart, Spence, Tony, Antiquity
The prodigious quantities of refuse recovered from excavations at Runnymede Bridge, Berkshire, England - and at other late prehistoric British sites - highlight those archaeological entities we call 'rubbish' and 'middens'. What is a 'midden'? General thoughts on an archaeology of refuse are applied to the specific case of these 1st-millennium BC sites in southern England in an attempt to comprehend their origin and scale in terms of the period's social geography.
The material of a midden is refuse or excreta. Refuse, comprising those objects and materials that have passed out of a sphere of active use and careful storage, is thus central to any understanding of a midden. However, we contend that it would be a mistake to regard archaeological concentrations of refuse as synonymous with middens.
The many transformations that can take place, both cultural and natural (Schiffer 1976), on a heavily used site such as a settlement mean that refuse can experience very variable histories (e.g. Bradley & Fulford 1980; Needham 1996). Similarly, the reasons for the aggregation of refuse may be very diverse. Schiffer's definition of four categories of refuse (TABLE 1) only begins to address the complexity of refuse movement and management, which collectively we refer to as the refuse-cycle [ILLUSTRATION FOR FIGURES 1 & 2 OMITTED] (cf. 'waste stream' Schiffer 1987). An object's refuse-cycle ceases with incorporation or burial, but may resume if it comes to be disturbed. Clearly we should be wary of supposing that even a 'coherent' context group would contain a homogeneous sub-assemblage in terms of its refuse history. Indeed Sullivan (1978: 201) felt that application of Schiffer's classification would tend to obscure the diversity of formation, or 'site building' processes.
Refuse has become a widely appreciated category of archaeological evidence in its own right, rich in significance for many aspects of social organization. Work has focused on a variety of aspects: discard behaviour and the redistribution of the material on and around sites of different character and economy (e.g. Binford 1983: 189-90; Deal 1985; Haydon & Cannon 1983; Murray 1980); studies of modern refuse disposal in the western world (e.g. Rathje 1974; Rathje & Murphy 1993); and taphonomic approaches to archaeological material, with detailed assessment of object condition (e.g. Sullivan 1978; Sorensen 1996; Serjeantson 1991; 1996; Rogers & Widdowson 1996). These last approaches are, of course essential, along with sediment studies, to the valid interpretation of deposit formation, and hence the character, organization and history of a site (Schiffer 1987). Our enquiry into the nature and meaning of middens stems from such an approach.
The use of 'midden'
The term 'midden' bears a range of connotations arising from contemporary western practices, ethnographic situations or archaeological contexts. We consider here some cross-cutting criteria germane to the definition and characterization of 'midden', going beyond the dictionary definition 'dunghill, manure-heap, refuse-heap' (Oxford English Dictionary (19048).
* Is it defined by its mode of construction - a particular way in which it came into being and was augmented?
* Is it to be of a certain size, to have some peripheral delimitation, and to have internal structure?
* Is it determined by its constituents - the types, the materials or their condition? Is consistency in these respects important?
* What about its role as a resource - is it essential that a midden served as a resource of some kind?
* Alternatively, is the critical aspect to be its function as a very particular mode of disposal?
* What importance do we attach to its spatial relationship to residences and the settlement at large? Does it matter what distance was involved and who was making use of it?
* Are temporal factors important - are there to be any limits in terms of speed or regularity of formation? …