Rock art research has changed in the last two decades, shifting from its early emphasis on empirical description towards increasing concern with interpretation and explanation. Conkey (1997) has linked this change to the use of ethnographic data and analyses, and nowhere is this better seen than in western North America. Although once argued to lack relevant ethnography (for example Heizer & Baumhoff 1962; Grant 1968; Steward 1968), recent North American studies have been largely based on syntheses of the ethnographic data that earlier archaeologists claimed did not exist (Whitley 2000; Keyser & Klassen 2001; Francis & Loendorf 2002; Keyser 2004; Sundstrom 2004).

This historical circumstance is more than merely anecdotal, however, because of its implications for the interpretive process. Most notable in this regard are the circumstances involving Robert Heizer, senior author of the most widely cited North American rock art monograph: The Rock Art of Nevada and Eastern California (Heizer & Baumhoff 1962). Although Heizer was a preeminent ‘dirt archaeologist’ renowned for his excavations in California, the Great Basin and Mesoamerica, he was also an accomplished ethnohistorian who was quick to recognise, and use, ethnographic sources for archaeological interpretation. (Indeed, the American Society for Ethnohistory in 1980 created a Robert F. Heizer Award, presented annually for the best article on Native American ethnohistory.) Perhaps more to the point, Heizer's first rock art publication, in 1953, was an ethnographic synthesis of California pit-andgroove engravings. His assertion that there was no rock art ethnography needs then to be qualified as a reference to no ethnography for paintings and engravings. Yet it also needs to be understood as a conclusion derived not from a failure to examine the ethnographic record, but instead almost certainly as a result of reading the major sources. Why did Heizer miss the ethnographic evidence?

I have suggested previously that this resulted from a literalist perspective confusing ethnographic commentary (which rightly should be understood as ‘raw data’) with complete exegesis (Whitley 1994a): failing exegesis amenable to literal western understanding, Heizer and others inferred that no meaningful data existed.