for some archaeological purposes; the question is whether or not there can also be “natural” (i.e., essentialist) types that have objective existence independent of any purpose of the typologist. The question in simplest terms is, Can types exist independent of purposes?

Both the erstwhile functionalists of the mid-century era and the new archaeologists of the next generation attempted to answer this question in the affirmative. It is evident, however, that the conception of what constitutes “naturalness” was quite different in the two cases. For the functionalists, it was a wholly emic concept, identifying that which was meaningful to the artifact makers and users. For the new archaeologists, despite their occasional assertions to the contrary, it was basically etic, identifying that which was discoverable through statistical procedures. There were contemporary critics of both approaches who continued to insist that types cannot be wholly separated from the archaeologists’ cognition, which in turn is shaped by his or her purposes. It remains true in any case that nearly all of the artifact types that are in general everyday use were developed for avowed culture-historical purposes while it has so far been possible to develop purely formal and purpose-free classifications only in the case of specific, limited assemblages.

Another issue raised in the typological debate was that of emic versus etic classification, although those two terms were not generally employed during the earlier phases of the debate. The mid-century functionalists insisted that the artifact types previously in use were somehow “wrong” because they were purely heuristic constructs while the proper aim of archaeology should be to understand (and classify) from the perspective of those who were studied. As applied to artifact classification, however, this approach proved to be less revolutionary than some of its proponents believed. It did not really affect the way that individual types were defined, only the way in which they were grouped into larger taxa.

Eventually, it was recognized that the categories created by the functionalists were ultimately those that responded to the archaeologists’ own logic and that the actual perspective of the prehistoric artifact user was beyond recovery. As a result, the functional approach, as an alternative to formal classification, died out after a generation or so, although it continues to be widely employed as a way of grouping artifacts for publication in monographs. It is noteworthy, however, that postmodernists are now raising the same questions about formal typologies that the functionalists raised, that is, casting doubt on the typologies because they reflect the categorical logic of the archaeologists’ own culture. But while the functionalists had believed that the prehistoric native’s outlook should be privileged, if only it could be recovered, the postmodernists now insist that nothing should be privileged.

Yet another issue raised in the debate was that of lumping versus splitting. Although it had long been recognized that some classifiers were naturally “lumpers” while others were “splitters,” the difference became a methodological issue with the introduction of statistical procedures in the late 1950s and even more with the advent of widespread computer use a generation later. Types had traditionally been devised by a partitioning or splitting procedure, dividing up a collection of material into recurring and visibly distinct types. In this approach, types were recognized first, and their specific defining attributes were then discovered through rigorous analysis. On the other hand, the method of attribute clustering is necessarily a lumping procedure in which attributes are defined first and are then clustered together to define types. There continue to be advocates of both approaches, as well as people who believe that these are alternate routes that will ultimately arrive at the same end. It remains true, however, that most of the artifact types that are in general use today were devised by the older splitting procedure.

The typological debate has been conducted mainly by North American prehistorians, particularly as regards the use of computers in classification. Since the 1960s, there has also been a somewhat different typological debate among European prehistorians, a debate that emphasizes methods rather than objectives. In the Old World, the influence of de Mortillet and Montelius