the American example was critically important to European archaeology of the same vintage (e.g., Jensen 1975; Lubbock 1868). Because the initial approach to antiquities generally was functional—an unremarked consequence of common sense—spatial and temporal patterns were obscured, and early accounts were ad hoc (e.g., Haven 1856; Jones 1873; Thruston 1890).

One important feature to note in regard to this initial recognition of the archaeological record is that archaeological data were entirely anecdotal, that is, they were accumulated on a case-by-case basis, largely by accident and without design. Thus, prior to the closing years of the nineteenth century, what sites and monuments were known were largely the product of accidental discovery. Systematic efforts to locate and enumerate sites, almost exclusively mounds and other monuments, initially were spurred on by conservation interests (e.g., Houck 1908; McLean 1879). Archaeological survey organizations, usually organized at the state (or province) level for maintaining such rosters, were creations of the twentieth century (Skinner and Schrabisch 1913). With few exceptions, the organized surveys are of minor research significance, since what they recorded and why remained ad hoc even if the effort to locate sites was not. The twentieth century saw the phenomena deemed worthy of recording expanded to include excavatable (i.e., more or less dense) concentrations of artifacts in the absence of monuments (Smith [1890] 1910) and in recent years the find spots of small numbers of artifacts or even, as some now argue, single objects (Zeidler 1995, appendix 1). These changes have almost always come about through debates on how to define site or, to more accurately reflect the spirit of the debates, what sites are. The literature illustrates well the absence of theoretical justification for the decisions being made. Although archaeologists certainly have reasons for what they do, the argument is made in terms of what is “real,” with the result that what parts of the record were recorded, described, managed, or saved has been a matter of historical accident. Whole classes of phenomena (e.g., small burial mounds) have been virtually obliterated, and others (e.g., low-density usages) are on their way to the same fate. Furthermore, the record has often been conceived in ways that preclude explanation (e.g., presence-absence [the site notion] rather than variable density [see Dunnell 1992b, Dunnell and Dancey 1983, and Thomas 1974 for an elaboration of these ideas]).

The range of objects that qualify as artifacts has undergone a similar expansion. Initially, only the most obvious shaped objects were treated as worthy of description and/or preservation. Anything broken, unfinished, or believed to be either was discarded. As whole objects became more rare, broken but “identifiable” objects were included. A few enterprising students (e.g., Wyman 1875) did describe such unshaped objects as shells and bones from middens before the turn of the twentieth century, but by and large such analyses and the systematic collections on which they were based came after the mid-twentieth century. Similarly, although the artificial nature of lithic debitage was recognized long ago, it was as late as the 1970s before its systematic collection and analysis became de rigeur. In some places (e.g., in New Zealand—see Shawcross [1964]) archaeologists moved more quickly in this direction without any fundamental change in methodology simply because of a paucity of shaped objects, but in the United States this shift had to await more fundamental change.

One element of U.S. archaeology, indeed of archaeology everywhere, that emerged early on was the recognition of the importance of conserving the archaeological record. The archaeological record was clearly fragile and being destroyed by vandals, the curious, and simple inattention (Atwater 1833; Schoolcraft 1847; see also Lubbock 1868; Worsaae 1848). As Fiske ([1815] 1820) put it succinctly nearly 200 years ago: The archaeological record is being lost because it is being found. This concern was, as just noted, the impetus behind the gradual shift to a systematic recording of monuments. As portable objects became more important, this same concern also led to the development of museums (in contrast to and at the expense of the “curiosity cabinets”). It is therefore not surprising