INTRODUCTION
The use of ethnography is probably the single most important interpretative method currently available to rock art researchers. In South Africa, almost all recent insights into San rock art are based on ethnography (for example Vinnicombe 1976; Lewis-Williams 1981; Lewis-Williams & Dowson 1999; Lewis- Williams & Pearce 2004; Eastwood & Eastwood 2006). This method, originally developed in South Africa for the study of San rock art, has now also been successfully transferred to other southern African rock art traditions (such as Cheŵa: Smith 2001; Zubieta 2006; Northern Sotho: Smith & Van Schalkwyk 2002; Moodley 2008; Van Schalkwyk & Smith 2004; Namono & Eastwood 2005). Globally, similar methods have also proved successful, particularly in the United States (Whitley 2000; Loubser 2006). Considering the success of the method, it is very likely that it can be much more widely applied, particularly in regions with relatively recent rock art and ethnographic accounts, such as northern Europe, Australia, South America and parts of Asia. Not to use available ethnographies in these contexts would be disingenuous and intellectually dishonest.
In this paper I examine various issues related to the use of ethnography in archaeology, particularly in the study of rock art. As such, the paper is primarily methodological, as opposed to theoretical, technical or, indeed, empirical.
The issue I discuss is not whether researchers can use ethnography to interpret rock art – this is a stale debate, and long demonstrated to be the case – but how and if one may use ethnography in cases where archaeological societies are demonstrably different from ethnographic ones, instances in which the contingencies of history have intervened. I use San ethnography and the Holocene Later Stone Age archaeology of the southern Cape of South Africa as examples.
In addressing this issue, I begin with a general discussion of analogy and how it may be used in archaeology. Although analogy is a topic that has frequently been discussed by archaeologists, I believe it is valuable to re-state the basic principles here because all uses of ethnography in archaeology are analogical, and the points I make ultimately rest on the logic of analogy. I then provide a brief overview of the sources and nature of San ethnography before discussing the ways in which it may be used to interpret early and mid-Holocene archaeological remains.