The role of narrative in explanation has received considerable attention in most of the disciplines concerned with questions of historical process, including history, geology, psychoanalysis and palaeo-anthropology. Archaeologists, however, have been curiously reluctant to consider the proposition that their reconstructions of the past are fundamentally narrative in character. An argument is put forward for the serious study of narrative in archaeology, and three case studies from the prehistory of the New Guinea Highlands are presented in support: a brief review of the debate over the impact of sweet potato on Highland society; an analysis of the changing interpretations of the Kuk Swamp agricultural site by Jack Golson; and a summary of the role of indigenous narratives in accounting for the history of wetland drainage amongst Huli speakers in the Southern Highlands of Papua New Guinea.
This paper considers the ways in which Jack Golson and other researchers working in the New Guinea Highlands have sought to explain the past--how they have conceived of the relationship between archaeological data and explanation, and how they have achieved consensus at different junctures over the past 35 years. My working assumption is that archaeological practice already contains 'a significant wisdom about the nature of archaeological reasoning' (Ballard 1995a:212). The challenge is not to trawl through other disciplines in search of better theories with which to tackle the problem of explanation in archaeology but rather to understand more closely the structure and nature of existing forms of explanation. Archaeological explanations evidently work, insofar as the discipline is able to proceed with shared understandings of plausibility, but it is not clear that archaeologists necessarily understand how of why their own explanations succeed in convincing their colleagues. What I seek to explore in this paper is the contention that archaeological explanation develops within and through frameworks that are fundamentally narrative in character. Furthermore, recognition of the role of narrative in archaeological explanation, far from undermining archaeology's pretensions to the status of a science, may allow us to generate explanations which are more fertile in terms of the questions that they raise and the spaces that they open up for further enquiry. To realise this potential, we need first to describe the way in which narrative functions in archaeological explanation. As the philosopher Paul Roth observes, 'without a theory of narrative explanations, there exists only an unanalyzed practice, a habit tolerated but not at all understood' (1989:453).
Another, more pragmatic consideration has to do with the current state of archaeological research in the New Guinea Highlands, where difficulties of access have seen a dramatic decline in active research since the heyday of Golson's work at Kuk and elsewhere during the 1970s and 1980s. In the absence of regular infusions of fresh data from new excavations and surveys, historians of the Highlands will increasingly have to develop novel and more subtle ways of interrogating and combining existing materials from archaeology and related disciplines; refining our analyses and interpretations in a return, perhaps, to the period during the early 1960s when researchers such as James Watson, and Susan and Ralph Bulmer speculated about the pre-colonial past on the basis of little or no archaeological evidence (see discussion of Watson below, S. and R. Bulmer 1964). What I want to suggest is that this need not necessarily be a bad thing. This last point essentially prefigures the general argument of this paper which is that archaeology, like other historical disciplines, has never been reliant on fresh material, to the extent that certain of its practitioners might have us believe, in order to advance its understanding of the past; advances in our understanding of the past can, and invariably do proceed in the absence of new material. …