Journal of Archaeological Research, Vol 6, No. 4, 1998
Status and Role of Formation Theory in Contemporary Archaeological Practice Michael J. Shott1
Since Binford appropriated the term "middle-range theory," it has signified the process of reasoning from the extant material record to the cultural past, Merton's sociological concept of middle-range theory is relevant to archaeology, but does not mean what Binford denoted by it. More accurately, Binford's domain should be called formation theory." By whatever name used, archaeologists differ greatly in our views of its role and status. Somehow, formation theory has come to be viewed as method but not theory, and as intrinsic to materialism, but irrelevant if not antithetical to other ontologies. Yet it is as critical to the contextual understanding of the past sought by many archaeologists today—a role that, among others, belies formation theory's marginal status in academic practice. KEY WORDS: middle-range theory; assemblage; formation; material record.
INTRODUCTION The material record is archaeology's mixed blessing, at once registering cultural meaning and process on a vast time scale and posing daunting interpretive problems. Archaeologists struggle constantly with the familiar problematics of inference because what we observe in the record does not register directly what we wish to know about the past. Instead, the two are mediated by assemblage formation. We have long acknowledged this mediation, if we have only indifferently regarded it. In response to this condition but from different perspectives, Binford (1977, 1981a, 1982) and Raab and Goodyear (1984; Goodyear et al., 1978) 1Department of
Sociology, Anthropology and Criminology, University of Northern Iowa, Cedar Falls, Iowa 50614-0513. e-mail:
[email protected]. 299 1059-0161/98/1200-0299$15.00 C 1998 Plenum Publishing Corporation
300
Shott
introduced the sociological concept of middle-range theory to archaeology. Schiffer (1988) later synthesized and critiqued the sources. Despite their efforts, uncertainty persists today about what middle-range theory is and how it figures in archaeological research. No short essay can settle the linked questions of middle-range theory's status and role, but this one assays the task. It synthesizes Merton, the sociological source of the concept, compares his meaning of middle-range theory with that proposed by Raab and Goodyear and Binford, who appropriated the term for use in archaeology, and identifies important differences between the sources. Binford (1983, pp. 18-19), in particular, recognized the differences between his use and Merton's of middle-range theory. These differences bear emphasis in view of the frequent if sometimes indiscriminate use of the term in archaeology. But at issue is much more than mere semantics. Whatever we choose to call it, what most archaeologists signify by middle-range theory remains critical to archaeological reasoning. Therefore, this essay defends this body of thought as an independent domain of theory significant in archaeology.
STATUS OF MIDDLE-RANGE THEORY Modern interest in middle-range theory arose in the intellectual ferment that characterized American archaeology of the 1960s and 1970s, exemplified particularly, of course, in the work of Binford and Schiffer. Binford (e.g., 1977, 1983) carried out ethnoarchaeological work to disentangle the complexities and contextual particulars of the formation of stone and especially bone assemblages, with the goal of revealing the systematic cultural practices that produced them. From a broader and more purely theoretical perspective, Schiffer (1976, 1987) documented the complexities of assemblage formation processes in general. Despite the many differences between them—well aired in print—Binford and Schiffer shared a concern for the validity of the inferences we make from the material record and an interest in how that record formed. Their views are not nearly as different as often assumed. Yet Binford and Schiffer are separated by some genuine differences in method, emphasis, and perspective as well as semantics. One of the apparent semantic differences was created by Binford's use of the term middle-range theory. Most archaeologists are at least passingly familiar with middle-range theory, yet their views on its importance and development vary widely. Some rank it among the most vital of disciplinary concerns, considering progress in its development as one of our most urgent tasks (e.g., Bettinger, 1991, pp. v-vii; Thomas, 1986). Many, though, consider the importance of
Status and Role of Formation Theory
301
middle-range theory overstated, to put it charitably. Some regard it as badly underdeveloped (e.g., Binford, 1981a, 1983; Cowgill, 1993; Kelley and Hanen, 1988; Schiffer, 1987, 1988); others consider it so overdeveloped as to overshadow our ultimate goal of understanding the past (e.g., Bradley, 1993; Moore and Keene, 1983; Wobst, 1990). Some consider middle-range theory critical to archaeology's theoretical goals, but others view its emergence as processualism's tacit admission to failure in attaining such goals (Yellen, 1989, p. 107), a sort of Merlinesque smoke screen intended to delude detractors or perhaps even practitioners. Wylie (1989, p. 96) condemned it as largely a failure in its own right, let alone as a mask to failure in other respects. Whatever its merits, to most this body of thought is theory (e.g., Bell, 1994, pp. 15-20; Bettinger, 1991, pp. 77-78; Binford, 1977; Kosso, 1991; Mills, B., 1994; Schiffer, 1988; Trigger, 1989), but to some it is method (e.g., Hodder, 1986; Raab and Goodyear, 1984), and to others it is irrelevant—no matter its status—to an archaeology that holds truly scientific aspirations (Dunnell, 1992, p. 87). Many archaeologists simply ignore the subject, perhaps because they are unconvinced of its importance or wary of its implications for the validity of their interpretations of the past. Our collective disarray about middle-range theory has several causes. One is the semantic confusion noted above, caused by attaching different meanings to the same term. Even though an obsession with semantics risks deterioration into sterile exercises in hair-splitting, any discipline requires reasonably clear and widely shared meanings for its important terms, which justifies this concern. A more important reason is philosophical, involving the importance that archaeologists attach to interpreting the material record through the agency of middle-range or equivalent theory by another name. As Schiffer (1988, p. 461) noted, theory itself is a slippery term, let alone the narrower middle-range theory. Here, theory means empirically grounded statements of relationship between variables useful, in the case of archaeology, to explain the cultural past. It is general if not universal in scope, not bound to specific contexts. Theory must be subject to proof, within the limits of practice set by existing knowledge. This definition resembles Schiffer's (1988, p. 463) experimental law; Schiffer's "theory" (1988, p. 462) might more accurately be rendered as "ontology." "Method" denotes simply how we gather and validate the evidence that is relevant to applying theory. What is theory and what is method, to adumbrate what follows, are determined by context, and theory in one may be method in another.
302
Shott
MERTON'S VERSION OF MIDDLE-RANGE THEORY The place to begin consideration of middle-range theory is with its originator in sociology, Robert Merton. Merton's earliest writing on the subject was a short comment appended to a paper by the eminent sociologist Talcott Parsons (1948), who called for more self-conscious theorizing by sociologists. Merton acknowledged the importance of general theory, but argued the equal importance of being able to test it against empirical data. In terms archaeologists have used since the 1960s, Merton identified general theory approximately with Kuhn's concept of the scientific paradigm and argued, as Kuhn and many others had, that it was not directly testable. Another sociologist, C. Wright Mills (1959), called it "grand theory" and argued against it on similar grounds. By whatever name we call it, making general theory susceptible to testing against empirical observation required an intermediate body of theory that was itself directly testable, theory that simultaneously embodied abstraction and groundedness (Turner, 1986, p. 88). This latter is what Merton later would call middlerange theory, defining it as theories that lie between the minor but necessary working hypotheses that evolve in abundance during day-to-day research and the all-inclusive systematic efforts to develop a unified theory that will explain all the observed uniformities of social behavior, social organization, and social change. (1967, p. 39) Middle-range theory, then, is what relates observation to paradigm, ontology, or philosophy. Merton's brief 1948 comment and his renowned later work (1957) both discussed middle-range theory, but his authoritative account emerged only in 1967. There Merton elaborated his original views, using some of the same points and indeed the same passages found in his earlier writings. Somewhere en route to the 1967 book Merton coined the expression middle-range theory itself. He identified the following as its key properties. 1. What It Is Not. Middle-range theory is not general theory. The latter is "a total system of sociological thought" (Merton, 1967, p. 45) or "general orientations toward data . . .rather than clear verified statements of relationships between specified variables" (1948, p. 165). A discipline needs such general theory, which nevertheless is inadequate by itself for scientific purposes. In isolation, Merton argued, general theories possess "architectonic splendor [but] scientific sterility" (1967, p. 51). Mills was more blunt. To him, general theory stood on "useless heights" (C. Mills, 1959, p. 33). Breaching this isolation requires middle-range theory. 2. Theoretical Status. This point may seem banal but it, too, bears emphasizing. Middle-range theory is not just empirical generalization, not the mere nose-counting (Merton, 1967, p. 53) derided by its critics. Instead, it
Status and Role of Formation Theory
303
is intrinsically theoretical; Merton defines theory rather vaguely as "logically interconnected conceptions" (1957, p. 5) from which "empirical generalizations have themselves been derived" (1967, p. 41). Similarly, C. Mills (1959, p. 73) advocated "an open and clear shuttle between broader expositions and more detailed information." 3. Independence of General Theory. Middle-range theory serves general theory by testing the latter against the empirical record. But specific middle-range theories must be epistemologically independent of specific general theories (Merton, 1967, pp. 41-43). They might be compatible with several general theories, although any general theories that shared many warranted middle-range components would be difficult to distinguish as a practical matter. 4. Sociologically (or, for That Matter, Anthropologically) Substantive. General theory concerns the explanation of "behavior, organization, and change" (Merton, 1967, p. 39). Significantly so, too, does middle-range theory, which can "account for limited aspects of social behavior, organization, and change" (1967, p. 51; cf. Raab and Goodyear, 1984, pp. 255, 257). Merton, in fact, repeatedly invoked "behavior, organization, change," making of the expression a virtual mantra in his discussion of middle-range theory. He identified middle-range theory with the study of class dynamics and conflict (1948, p. 166), of deviant behavior (1967, p. 51), and other theories of action and structure. Indeed, one example of what Merton (1967, p. 63) meant by middle-range theory was Weber's classic, The Protestant Ethic and the Spirit of Capitalism. Merton's (1938) own famous thesis on the Calvinist origins of modern science also might be considered a good example of middle-range theory. Few would doubt the substantive nature of such scholarship. Clearly, Merton did not conceive of middle-range theory as the theory or least of all the method of data collection, reconstruction, or analysis, important as these acts are. Instead, it is a theory of substantive phenomena, of human behavior in its cultural and social context. Merton's middle-range theory, in short, is substantive behavioral theory rather than theory of how to infer structure or behavior. 5. Testability. Middle-range theory is merely a link in a longer chain of inference that runs from general theory to observation. To Merton (1967, p. 51), as to many scholars and philosophers of science, it must be susceptible to verification. Otherwise it reduces to "formal and cloudy obscurantism" (Mills, 1959, p. 75). Precision—in this context meaning fidelity to empirical data—is vital to testability (Merton, 1957, p. 98). If middle-range theory is not testable then general theory, no matter its "architectonic splendor," reduces to metaphysics.
304
Shott
6. Critical to Scientific Progress. In the current intellectual climate it may sound naive to speak of science and progress; it is the sort of talk that prompts knowing glances and wry smiles. It also is worth the risk. Commitment to a scientific view of the past does not entail a particular ontological or theoretical position (Peebles, 1992). Merton (1957, pp. 7-10; 1967, pp. 50-53) clearly saw middle-range theory as an instrument to promote the advancement of knowledge—of general theory—in his field: "sociology will advance insofar as its major (but not exclusive) concern is with developing theories of the middle range, and it will be retarded if its primary attention is focused on developing total sociological systems" (1967, pp. 50-51). From a different perspective, Mills (1959, pp. 76-99) endorsed a similar integration of theory with observation and action. Both sociologists advocated a balance of interests between data collection and theory development at the middle and general levels. Neither wished middle-range theory to dominate sociology to the exclusion of general theory; on the contrary, both considered middle-range theory a critical instrument in evaluating general theory. To Merton and Mills, middlerange theory and general theory were complementary, not competing. Whether or not sociologists still use the terms or the domains they signify, middle-range and general theory remain pertinent in archaeology.
MIDDLE-RANGE THEORY IN ARCHAEOLOGY In the 1970s Binford appropriated the expression middle-range theory for archaeology. Binford (1983, pp. 18-19, fn5) explained how he arrived at use of the term, demonstrating in the process his awareness of sociological middle-range theory and arguing for a different sense of the term in archaeology. From a different route, Raab and Goodyear (1984; see also Goodyear et al. 1978, pp. 161-162) simultaneously introduced middle-range theory to archaeology. Their 1984 paper may derive from an earlier, unpublished one that antedated Binford's first use of the term, but Raab and Goodyear's 1984 paper was their sole extended publication on the subject. Their earliest publication (Goodyear et al., 1978) appeared a year after Binford's earliest one and cited Binford among its important sources. Binford's first extended use of the term middle-range theory was in his 1977 essay. There he emphasized archaeology's ultimate concern with prehistoric cultures, how they were constituted, and how and why they changed through time. Understanding, though, that the material record cannot be read as a straightforward, let alone nearly complete, account of past cultures and their transformations, Binford advocated development of a body of theory to be applied strictly to the material record. This theory
Status and Role of Formation Theory
305
would be, ipso facto, uniquely archaeological because only archaeology studies and interprets the archaeological record. The body of theory rested on the difference between the fixed nature of the existing record that archaeologists observe directly and the protean cultural processes and practices that created it. Middle-range theory in effect would link these two domains, attaching meaning to the material record in terms of the processes that created it (Fritz, 1972, p. 140). Binford argued that middle-range theory provided accurate means of identification, and good instruments for measuring specified properties of past cultural systems . . .we are looking for 'Rosetta stones' that permit the accurate conversion from observation on statics [the material record] to statements about dynamics [the cultural past]. (Binford, 1981a, p. 25)
So conceived, middle-range theory is an archaeological version of methodological uniformitarianism (Bailey, 1983, p. 175; Fritz, 1972, p. 140; Gould, 1965). Actualistic studies, experiments, and ethnoarchaeology would be important sources of middle-range theory. This belief inspired Binford's own ethnoarchaeological research. Indeed, Binford (1981a, p. 29; see also Grayson, 1986, pp. 77-78; Trigger, 1989, p. 362) practically defined middle-range theory in terms of ethnoarchaeological research. Today most archaeologists regard the latter as an important, but not the sole, source of middle-range theory. Binford (1981a, p. 32) himself identified historical (see also Leone and Crosby, 1987) and experimental controls for the conduct of middle-range research in addition to ethnoarchaeology. Both Binford's and Raab and Goodyear's views on the independence of general and middle-range theory were somewhat ambiguous, though similar to Merton's when closely examined [see Tschauner (1996, pp. 2023), who captured as well as reflected this ambiguity]. Just as Merton acknowledged the need for concurrent progress in both domains—"the development of general and middle-range theory must proceed hand in hand" (Binford, 1977, p. 7; see also Eggert, 1982, p. 145; Fritz, 1972; Kelley and Hanen, 1988, p. 289; Mills, 1994, pp. 56-57)—Binford seemed originally to imply that middle-range work must occur within some particular theoretical context. Otherwise, he argued, we may produce middle-range knowledge that is irrelevant to general theory (Binford, 1977, 1983, p. 50). In this view, middle-range theory serves particular general theory, not general theory at large, and in that sense is dependent upon the particular general theory. This certainly is the received postprocessualist view. Perhaps Binford meant that such theory must be relevant to general theory; relevance does not entail logical dependence (Fritz, 1972). Indeed, he (Binford, 1981a, p. 29) later stressed more strongly the independence of middle-range and general theory. Nevertheless, his ambiguity invited the
306
Shott
conclusion that Binford did in fact link middle-range theory with a particular body of general theory, broadly materialist in his case (Wylie, 1989). Binford also believed that middle-range theory was badly underdeveloped in archaeology. His 1977 paper essentially was a clarion call advocating the reliable interpretation of the archaeological record. It has been heeded by some who read Binford and, from a somewhat different perspective, Schiffer (1976, 1987), but Binford later could still conclude legitimately that "An observational language [i.e., middle-range theory] is essentially nonexistent in archaeology" (1981a, p. 24).
DIFFERENCE BETWEEN MERTON'S AND BINFORD'S MIDDLE-RANGE THEORY
Merton and Binford both described significant theoretical domains in their respective fields, and both called these domains middle-range theory. Moreover, Binford's theory shared most of the key properties of Merton's middle-range theory. Both are testable, independent of general theory, and important to progress in their respective fields. Merton's and Binford's views were similar in other ways as well because both were advanced largely in reaction to opposing views of prominent disciplinary colleagues. Merton reacted to Parsons' grand social theory, which he evidently regarded as "architectonic" (clearly not a compliment in Merton's lexicon) if not metaphysical. Binford reacted broadly against the cultural-historical empiricism that dominated American archaeology at the start of his career, and more specifically against Bordes' normative argument regarding French Mousterian industries (Shott, 1996a). But Binford's middle-range theory differs from Merton's in one key respect: it is not a body of substantive theory of cultural or social behavior, organization or change. It undeniably is critical to archaeology, because it is the theory that links what is at hand—the material record—to what we wish to observe, understand, and explain—past behavior and organization in its cultural context. Merton took for granted sociologists' ability to observe and measure directly that in which they held an interest. Archaeologists, in contrast, must infer what interests them from the static material record observed at first hand. That, substantially, is the function of Binford's middle-range theory, and its meaning was clear enough to a cultural anthropologist: "[middle-range] theories are not meant to be theories of behavior, predicting what people will do under particular material (i.e., physical environmental) circumstances. They are rather meant to predict the material consequences—in terms of the formation of archaeological
Status and Role of Formation Theory
307
sites—of particular circumstances of behavior" (Ingold, 1992, p. 798; see also Bailey, 1983, pp. 174-180; Salmon, 1982, p. 170). Thus Merton's middle-range theory begins where Binford's ends; they are not the same thing. Binford (1983, p. 19 fn5, p. 433) acknowledged as much himself in explaining the origin of the term as he used it and as Merton meant it. Raab and Goodyear's (1984; see also Salmon, 1982, p. 170; Schiffer, 1988, p. 462) criticism is valid; Binford did coopt the term, giving it a different meaning.
RECONCILING MERTON AND BINFORD
The confusion about meaning is exemplified in Bettinger's (1991) attempt to reconcile Merton's and Binford's concepts of middle-range theory. Amidst the semantic clutter, Bettinger coined yet another neologism; though citing middle-range theory, Bettinger referred more often to "theories of limited sets." It is the job of limited theories to reconcile general principles to particular cases by showing how such cases result from the general principle in the presence of special conditions .... It is, further, the interaction between the general principle and the salient properties of the special, set-defining conditions that accounts for what is observed. (Bettinger, 1991, pp. v-vi)
This is a lucid definition of Mertonian middle-range theory. Yet Bettinger (1991, p. 62) then spoke of "the general definition of middle-range research as an enterprise devoted to the assigning of meaning to empirical observations about the archaeological record." This is Binfordian middle-range theory. The source of Bettinger's (1991, p. 80) ambivalence is clear: "I see no qualitative, that is epistemological, difference between what have been called . . .middle-range theory (sensu Merton and Raab and Goodyear) and theories of the processes governing site formation (sensu Binford)." [See also Bell (1994, p. 16), who considered Merton's and Binford's concepts to be "distinctions without a difference."] If, as Bettinger argued, there is no substantive difference in these understandings of middle-range theory (as noted above, neither Binford nor Schiffer shared this view), then Merton and Binford have equal claim to middle-range theory. But Bettinger (1991) usually equated it with his "theories of limited sets," which themselves clearly are substantively cultural (1991, pp. 64-77, p. 107). As above, this is Merton's middle-range theory alone; it cannot also be Binford's.
308
Shott
A NEW LABEL FOR BINFORD'S MIDDLE-RANGE THEORY
It is not enough merely to argue that we should retain middle-range theory in Binford's sense because it has grown entrenched in the field. It is common, if not entrenched, but we have established no clear consensus on its meaning; instead we continue to argue about what it is. More importantly, the term has a relatively clear meaning, established earlier in another social science, and that meaning is not the same as Binford's. Mertonian middle-range theory is as important for archaeology as it is for sociology, a point perhaps overlooked in the running argument about Binford's version. Thus we should use the expression "middle-range theory" but we should apply it to the archaeological equivalent of Morton's sociological middle-range theory. Binford's theory of behavioral or organizational inference, not explanation, from the material record is equally important for archaeology. Obviously we must maintain an active interest in this theoretical domain, but to avoid further confusion we should no longer give it Merton's label. Binford himself made contributions to substantively middle-range theory (sensu Merton) in archaeology (Raab and Goodyear, 1984, p. 264; cf. Schiffer, 1988, p. 463). True middle-range theory is any theoretical explanation of observed behavior, organization, or structure that can be subsumed by more than one paradigm or philosophical approach. Thus Binford's (1980) well-known construct of logistic and residential foragers is an example of Mertonian middle-range theory (Bettinger, 1991, p. 64). It is substantively cultural; it is at least partially testable; and it is conceptually independent of any particular paradigm, although it bears close affinities to the various anthropological forms of materialism. Its independence, however, is illustrated clearly in Ives' (1990, pp. 315-317) use in a structural, not materialist, model. So, too, has Schiffer (e.g., 1979) formulated genuine middle-range theory for archaeology. Optimal-foraging theory in its various forms is yet another example of true middle-range theory (Bettinger, 1991, p. 107; O'Connell, 1995), although it is closely identified with behavioral ecology and classical economics, themselves variants of materialism. But theories like agency and internal contradiction that have gained status within the field also are derived from particular theoretical or ontological positions. They are not a priori just for this reason, and they certainly can be middle-range theory. It follows that what Binford called middle-range theory must have a different name (cf. Thomas, 1986, p. 245). This is no place to assay the history of archaeology's concern with interpretation from the material record for which, fortunately, good sources exist (e.g., Trigger, 1989). But it
Status and Role of Formation Theory
309
is important to place Binford's view in the context of at least the recent past, which shows that his middle-range theory pertained to what other archaeologists have called different things. Here is the place to seek a more accurate term.
OTHER APPROACHES TO THE QUESTION
Cowgill (1970) discussed the daunting problems posed in reasoning from the material remains recovered, to those deposited in the past and eventually to the organization and behavior that produced the record. He (1970, pp. 162-163) described sampling concepts ranging from populations of behavior in the past, to the "physical consequences" left behind by those behaviors, to the population of "physical finds" that archaeologists then sample imperfectly. CowgilPs treatment emphasized the problematics of reasoning to the past from what refractory evidence of it exists today, but he proposed no name for this important endeavor, perhaps because archaeologists were just beginning to appreciate its importance. Shortly, however, that realization developed. In a celebrated paper, Clarke (1973) classified archaeological theory, in the broadest sense, in terms of five connected domains. Clarke called the first "predepositional and depositional theory," defined as relationships between specified hominid activities, social patterns and environmental facts, one with another and with the sample and traces which were at the time deposited in the archaeological record. (Clarke, 1973, p. 16)
The first part of this definition concerns the theory that governs activity in its cultural and social context, which, presumably, Clarke intended as the "predepositional" component of this domain. The second part, Clarke's "depositional theory," concerns how pattern and process in the past create the archaeological record. Substantially, this is the same thing denoted by middle-range theory. Yet Binford (1981a, p. 25, fn2) equated his middlerange theory with Clarke's (1973, p. 17) "interpretive theory." The latter bears some affinities to Binford's concept but is closer to Merton's version of middle-range theory because it describes the process of testing general theory against observation or inference. No more than Binford did Clarke consider depositional theory sufficient to an archaeology holding reasonable pretensions to status as a cultural discipline. It was critical but so were predepositional theory and the remaining domains described in his paper (Clarke, 1973, pp. 16-17). All were needed to rescue archaeology from a possible fate as, in Clarke's (1973, p. 16) memorable phrasing, "an irresponsible art form." Clarke's depositional theory indeed captures the nature
310
Shott
of the archaeological task—inference to past culture from material deposited in the record—yet calls to mind geology more than archaeology. Trigger (1989, p. 357) suggested the expression "archaeology as itself," but this would sound cryptic as a label for a major domain of archaeological inference. More succinctly but in the same vein, Raab and Goodyear (1984, p. 259) preferred archaeological theory for the domain claimed by Binford's middle-range theory (see also Fritz, 1972, p. 137). Even Binford and Binford (1968, p. 2) described as archeological theory what Binford later would define as middle-range theory. Binford explained his eventual disenchantment with "archaeological theory" (1983, pp. 18-19, fn5), shifting to middle-range theory. By then, he had equated archaeological theory with general theory (1982, p. 22). The simple archaeological theory—sensu the Binfords and Raab and Goodyear—is appealing because it distinguishes what we do as archaeologists from what we do subsequently in cultural interpretation. However, it might be confused too easily with everything— including broader cultural theories—that concerns archaeologists. For this reason only, it should not be used. Schiffer (1988) took a more comprehensive approach to the problem. Like Clarke, he (1988, p. 464) described several theoretical domains that include "reconstruction theory," which "permits human behavior and environmental conditions of the past to be ascertained." Schiffer (1988, p. 469) was at pains to emphasize, contra critics (Binford, 1981b, 1983, p. 221; Dunnell, 1992, p. 87), that the subjects of reconstruction are not just short, discrete episodes of behavior but the organization and structure of past cultures as well, although usually he emphasized the former. Reconstruction theory itself consisted of correlates (e.g., fornrfunction identities in stone tools or pottery), and cultural and natural formation processes (1988, p. 467). C-transforms—the "nomothetic study of cultural formation processes" (Schiffer, 1988, p. 471)—are perhaps the most important of these domains, involving inference based on the culturally determined relationship between systemic and archaeological context. But archaeologists do not reconstruct the past, which is gone. Instead, we infer its nature from the material record that we directly observe in the present. As Binford (e.g., 1981b, 1983, pp. 65, 391; see also Dunnell, 1992, p. 87) argued, the target of inference is not reconstructed ethnographic tableaux but the operation of cultures over longer time and space scales. Indeed, Schiffer himself (1988, p. 469) registered some dissatisfaction with "reconstruction"; the term implies that the record is a distortion of the past rather than its direct product. Moreover, Schiffer (1976, 1988) already had coined another term, "formation processes," that is more appropriate. In different ways depositional, reconstruction, and archaeological theory partly capture the essential archaeological, but not anthropological,
Status and Role of Formation Theory
311
task, and all avoid the flaw of inviting confusion with Mertonian middlerange theory. As above, however, each possesses compensating drawbacks. Yet archaeology already possesses a suitable term. We deal directly only with assemblages of materials and contexts that were created by formation processes. The theory that governs this domain should be identified as "assemblage-formation theory" or simply "formation theory" to denote our understanding of how the archaeological record formed. Formation theory avoids conflation with depositional theory in geology and cultural theory in anthropology. It also avoids the problematics of reconstruction noted above. Like Binford's middle-range theory, Schiffer's (1988, p. 471) c-transforms, together with n-transforms or taphonomy, substantially are formation theory. Correlates also were part of Schiffer's (1988, p. 464, Fig. 1) reconstruction theory. They specify how artifacts were made and used; correlates are what Grayson (1986) identified exclusively with middle-range theory. Correlates certainly are important, but they are a theory of artifacts, not the archaeological record and how it formed. Formation theory is.
RECONCILING BINFORD AND SCHIFFER Binford and Schiffer, in their different ways, long have been the chief advocates of formation theory in American archaeology. It puts it mildly to say that they waged an acrimonious debate over their respective views of the status and role of formation theory in archaeology. Fools rush in, the saying goes, and no one should attempt lightly to reconcile their views; there are genuine differences of emphasis and substance between Binford and Schiffer that defy reconciliation. The first difference is one of cultural emphasis (Tschauner, 1996, pp. 8-10). Binford always saw cultures as systems that operate over large temporal and spatial scales and that produce material remains reflecting those scales (see also Stern, 1994); Schiffer demonstrated more interest in reconstructing shorter-term contexts. This is the basis for Binford's (1981b; cf. Schiffer, 1985) familiar criticism of Schiffer's research as a quest for so many Pompeiis. Tschauner (1996, pp. 9-10) synthesized the views with the argument that Binford's larger scale consists of the elements of Schiffer's smaller one, although the difference can be as substantial as the difference of kind, not just degree, between a photograph and a moving picture of the same subject. Binford exaggerated Schiffer's obsession with metaphorical snapshots of the past, and Schiffer (e.g., 1979,1988, p. 469) legitimately could claim an interest in processes that unfolded over larger scales. Nevertheless, Binford was more persuasive in this respect: the archaeological record is not a simple register of timeless ethnographic tableaux. We re-
312
Shott
construct or infer the operation of cultures at larger scales that are not commensurate with the ethnographic record. Following from this view, Binford saw formation processes as the cultural system, or at least as integral components of it. Contra Grayson (1986) and Tschauner (1996, p. 8), Binford's formation theory comprehends Schiffer's transforms and correlates alike. Schiffer's view seems more nuanced than the simple statement that assemblage formation distorts culture's registration in the record (cf. Binford, 1981b), but this conclusion can be drawn from casual reading. Perhaps the most significant difference between Binford and Schiffer, however, concerns formation theory's logical relationship to general theory. Obviously, the two involve different kinds of things, processes and relationships; they differ substantively. Equally as obvious, Binford and Schiffer agreed that formation theory is only as good as its ability to test general theory against the material record. But Schiffer consistently regarded formation theory as logically, as well as substantively, independent of general anthropological theory. In contrast, Binford's position was more ambiguous. To him, formation theory must be developed and applied as part of models of particular cultures; their operation is derived from general theory. Thus Schiffer's formation theory is independent of general theory both in logic and in substance, Binford's theory is definitely independent in logic and less certainly in substance. Binford (1983, pp. 163, 165) criticized Schiffer's interest in formation theory in absolute terms because he could not imagine the operation of such processes in the absence of some conception of culture process derived from general theory. On balance, Schiffer was more persuasive in this respect: it may be useful but it is not critical to have a "target context" (Binford, 1983, p. 163) for the application of formation theory. Just as general theory can be developed without specific empirical referents, so, too, can formation theory. Their important similarities should not be obscured in the close attention devoted to the differences that separated Binford and Schiffer (Wylie, 1995). Both were concerned to validate inferences made from the material record; both were committed to some version of Wylie's (1989) "qualified objectivism." Both pursued unapologetically a scientific archaeology that does not deny the legitimacy of humanistic and other views of the past. FORMATION THEORY AS THEORY
Whatever we choose to call it, our understanding of assemblage formation processes requires theoretical as well as empirical grounding. Raab
Status and Role of Formation Theory
313
and Goodyear (1984) argued that sociological middle-range theory is not substantive theory of social behavior, organization, or change. This judgment is difficult to square with Merton's (1967, pp. 39-72) own account in which, again, theoretical substantiveness is integral. As noted above, Raab and Goodyear also advocated restoring the label "archaeological theory" to what Binford called middle-range theory. But following from their view of sociological middle-range theory, they classified archaeological theory as method [Raab and Goodyear, 1984, pp. 258-259; from a different perspective, see also Taylor (1948, p. 41)], which it most certainly is not (Kosso, 1991, p. 622; Mills, 1994, p. 56; Schiffer, 1988, p. 463). Any theory can be used as method depending on the context and purpose; radiocarbon dating is a textbook example in archaeology, in which the physical theory of isotopic decay is used to date materials of interest. Isotopic decay is no more a mere method than is the archaeological theory concerned with the systematic relationship between cultural behavior and material remains. In fact, archaeologists do use middle-range theory as method. Ultimately, our goal is to learn why the past unfolded as it did, for which we use general theory. It is not merely to learn what the past was like, for which we use formation theory. But theory commonly is used as method, and formation theory as method nevertheless remains a body of theory. It is both, depending on how it is used. Wylie (1989; see also Bailey, 1983, pp. 177-179, Fritz, 1972, pp. 137, 157) pointed out that Binford's principles of assemblage formation and the ethnoarchaeology on which much of it is based are theory dependent. This much is undeniable, if anyone ever attempted to deny it; it also is another way of saying that formation theory is just that, theory. Faced with similar interpretive challenges, paleontology recognizes a similar theory dependence (Kitts, 1992). Binford's formation processes are not purely objective but themselves dependent upon a "rich theoretical" (Wylie, 1989, p. 100) panoply. No less is true of formation processes in general. Thus Wylie's (1989, p. 105) brief is reasonable for qualified objectivism with respect to formation theory rather than dubious claims of absolute objectivity. Moreover, we do indeed require her nuanced assessment (1989, p. 108; see also Tschauner, 1996, p. 24), both of general theory and of the empirical and theoretical basis of assemblage formation. For instance, the correlation between size and use life in pottery vessels (Shott, 1996b) is, in one respect, merely an empirical generalization, perhaps a low-level (Wylie, 1989, p. 100) one. Size does not directly determine use life. Instead, it is a crude but useful proxy for a host of factors—symbolic value, the nature, frequency, and context of use, and structural and mechanical properties of temper, paste, and construction— that, in complex interaction, determine vessel use life. Theory of this nature
314
Shott
is testable. Indeed, the relationship of size and use life does not hold universally (Shott, 1996b), and in that sense theory is falsified in some contexts. As archaeologists continue to develop carefully the theory advocated by Wylie (1989, p. 108), they can determine its limits or replace it entirely with something better based on knowledge of other determinants of use life and the nature and magnitude of their effects. Until then, however, we can act as though size indeed determines use life and, in this strictly limited sense, treat the correlation of size and vessel use life as a theoretical proposition—an element of formation theory—capable of application to archaeological data. There are many good reasons to want to know or at least estimate vessel use life (Schiffer, 1987; Shott, 1996b), and they are consistent with many general theories, not just materialism. Use life can aid in estimating settlement occupation span (e.g., Hill, 1995; Mills, 1994), itself useful to the meticulous inference of regional occupational histories, and the role of political processes, individual and social agents, and economic factors in the establishment and eventual abandonment of those settlements (e.g., Nelson et al., 1994). It also can determine the size and composition of vessel inventories in use, which bears on activities and socially and economically determined variation in them (Hill, 1995; Lightfoot, 1994; Varien and Potter, 1997). Use life can influence where and why style registers in material culture (Braun, 1995, pp. 134-135), with important implications for how stylistic attributes are fixed and transmitted. Processualists, behavioralists, Marxists, selectionists, symbolists, and students of mind and agency alike should be interested in questions of this nature and many others that can be answered reliably only with the aid of formation theory. It follows that many archaeologists should be interested in estimating pottery vessel uselife. This much requires due regard for the value of formation theory to archaeological inference. Thus this element of formation theory is independent of but has broad relevance to general cultural theory. In the 1970s Binford and Schiffer appealed for research on formation theory, both warning that advance is likely to be slow in its immature state. In different ways, both have achieved significant progress. And formation theory continues to develop. Over a decade ago, for instance, Schiffer (1985) argued the importance of determining artifact use lives. He offered admittedly crude use life estimates (Schiffer, 1985, pp. 33-34), stressing at the same time their provisional quality and the need to develop more accurate and reliable ones. We continue to approach that goal (e.g., Shott, 1996b; Varien and Potter, 1997) with better estimates that nevertheless remain crude, just less so than before.
Status and Role of Formation Theory
315
INDEPENDENCE OF GENERAL AND FORMATION THEORY Most archaeologists agree that formation theory must remain independent of but develop concurrently with general theory. Nevertheless, archaeologists of different theoretical persuasions have challenged the independence of formation theory, either to attack or support it. In all such cases, however, the views are reminiscent of Hawkes' (1954) ladder of inference. As a brief for circumspection in inference from the record, Hawkes' ladder was salutary; too often, however, it served as a counsel of despair preventing archaeologists even from contemplating inference from the record that is simultaneously rigorous and culturally meaningful (Bailey, 1983, pp. 176-177). Hodder's position, for instance, seemed a priori more than reasoned (1986, pp. 16, 93,103,116): "There can [n]ever be universal laws of cultural [formation] process which are independent of one's higher-level cultural theories" (Hodder, 1986, p. 103; cf. Kosso, 1991; Tschauner, 1996). From a very different perspective, O'Connell (1995, p. 233) did not explain why formation theory must depend on his preferred general theory, evolutionary ecology. There is no reason why this general theory cannot be tested using an independent, not an adjunct, body of formation theory. Hewing closely to Hawkes, Trigger (1989, p. 394) saw formation theory as relevant mostly to materialist general theory, if not logically dependent upon materialism. Likewise in sociology, Turner (1986, pp. 88-89) argued that Merton's middle-range theory served principally Merton's brand of functionalism rather than sociological theory at large. Whatever the case in sociology, there is no doubting in archaeological practice that formation theory has done service almost exclusively to materialist approaches. Yet this need not be so in principle. Trigger (1989, pp. 395-396) argued that cultural behavior is too complex, diverse, and arbitrary with respect to material conditions to be inferred reliably by formation theory. Cultural behavior is mostly what Trigger said it is. Complexity and diversity, however, do not render cultural behavior unsystematic, just not reducible to material conditions as Trigger (1995, pp. 451-452) himself more recently argued. When and to what extent cultural behavior also is materially arbitrary remains poorly known but knowable. Otherwise, its refractory qualities are not insurmountable, merely daunting. Perhaps for similar reasons, Saitta (1992, p. 891) considered formation theory especially pertinent to supposedly simple cultures like bands but not to states. Like others, though, Saitta emphasized the independence of formation theory and general theory. If the latter reaches complex societies then so, too, must formation theory; otherwise, formation theory would in fact
316
Shott
depend upon that general theory that pertains to band societies, and we would lack comparable theory pertinent to complex societies by which to validate theoretical inferences. But this cannot be. Instead, we must develop what Cowgill (1993, p. 562; see also Trigger, 1995; Wylie, 1989, p. 101) called "middle-range theory of the mind," which yields validated generalizations, if not universals, of symbolic practice grounded in empirical observation that are applicable to societies at all degrees of complexity. Today as in the past, mind is a quality of foragers and farmers just as much as of anyone else. This plea is less for a distinct body of theory than for the extension of formation theory to the broadly cultural and symbolic from the narrowly material. Hill's (1995) recent study is a good illustration of this prospect. Wobst (1990) agreed with Merton and Binford that formation theory can be independent of general theory. Presumably he also would agree that formation theory and general theory can progress simultaneously. Merton felt that a discipline requires no commitment to particular bodies of middle-range theory before it addresses general theoretical questions. But Wobst argued that in practice formation theory in archaeology is used only to support—when not actually overshadowing—a particular general theory, broadly the materialism of processual archaeology. [Bettinger (1991, p. 61) held a similar view, but approvingly. In sociology, Mills (1959, pp. 48-49) attributed the same function to general theory.] Moreover, formation theory and the general theory it supports are believed to share a particular, covert ideological agenda in the guise of value-free science. In this way, Wobst (see also Moore and Keene, 1983, p. 6) saw formation theory as not merely a scientific but intrinsically a political instrument. It acts in deliberate support of, or unwitting duplicity in, an extrascientific political program. The charge is political naivete or worse. This charge is reasonable if arguable (Cowgill, 1993; Saitta, 1992; Wylie, 1989). Accepting its validity for argument's sake, there still is no necessary reason why formation theory should dominate general theory or why it could be used only in the service of one kind of such theory. In effect, the problem Wobst identified is one in practice, not in the nature of formation theory. This is demonstrated by the similar function of midcentury general theory in sociology (Mills, 1959). The problem cannot lie in their nature but in their conditions of use if general theory supports legitimation in one field and formation theory supposedly supports it in another in opposition to general theory. More serious still, Peebles (1992) identified formation theory with an exclusive interest in past behavior to the neglect of mind and symbol, and with a epistemological commitment to logical positivism. A necessary link with behavior is questionable, although Binford's (1982) most extreme statements against mind make the charge plausible. Peebles (1992, p. 360)
Status and Role of Formation Theory
317
argued that the behavior that formation theory infers is used to test tautological behavioral theory. It is not merely that formation theory is not independent of a particular body of theory in practice but that it cannot be in principle. Peebles apparently would deny the independence of formation theory and general theory as claimed by Merton and many archaeologists. In his view, formation theory is part of a larger package of behavioral theory and cannot function outside of that context. As its promoter, formation theory "is largely responsible for the poverty of much of contemporary archaeological theory" (Peebles, 1992, p. 359). Moore and Keene (1983, p. 4) expressed the same view nearly a decade earlier, speaking of "methodological exactitude" combined with "conceptual lethargy." Here, formation theory is not merely a symptom of the ills of current practice but a cause of them. Ironically, there is no poverty of theory but an undisciplined plethora of them—ontologies perhaps, predilections certainly—that clamor for attention and support in archaeology. A glance at the pages of any recent issue of major journals illustrates this condition, not in itself a bad thing. We are impoverished instead in the means to evaluate rigorously competing theories against the material record and so to determine which among them provide more satisfactory accounts (Shott, 1998). Theory untested is theory abused or no theory at all; Peebles's call for a richer array of theory requires a greater, not lesser, role for formation theory. Development of formation theory applied to relevant empirical observations, for instance, might as easily promote as impede Peebles's symbolic alternative to what he considered the pernicious effects of behavioralism. Because formation theory is independent of general theory, particular elements of formation theory can be encompassed by more than one body of general theory (cf. Hodder, 1986, p. 103, 1991, p. 36). Saitta (1992, p. 887) made the point clearly: "Radical archaeologies need . . .to develop middle-range frameworks [sensu Binford] relevant to their specific theoretical interests in social difference, tension, conflict and negotiation." Of course, archaeologists of various theoretical persuasions, not just Saitta's, may be interested in class and internal conflict in prehistoric cultures (Shott, 1993), so a distinct body of radical formation theory is neither necessary nor desirable. Indeed, some (Kosso, 1991; Tschauner, 1996) would go so far as to identify Binford's middle-range theory with Hodder's hermeneutics, not in the content of their respective theories but in their strategic goals of attaching cultural meaning to patterning in the material record. Hodder's earlier and least doctrinaire work invoked "general principles of symbolic meaning" (1982, p. 186) that sound much like CowgilPs "middle-range theory of mind" or symbol, and there is empirical warrant for such principles (Douglas, 1966) that archaeologists could examine. If
318
Shott
Binford and Hodder can, to some extent, be reconciled at the level of formation theory then so, too, can radical and other archaeologies. But Saitta made a fair point and, moreover, registered an interest in cultural and social theory closely akin to Merton's interest in middle-range theories of deviance and class conflict. Thus some have claimed but none has yet demonstrated how formation theory must be linked to particular general theories. There is nothing in the existing theory specifying, for example, the determinants of tool design or use lives that links these properties necessarily to any particular general theory. To control for the effects upon assemblage formation of design and use life requires no a priori commitment to a particular view in general theory. Formation theory permits reasonable inferences about things such as the time span over which settlements were occupied, the size of resident groups, and their structure and organization. It helps us to infer what items were used together and in what way, among the many that may have been deposited together. The deductions reached from the careful application of formation theory not only are accessible to archaeologists of all theoretical persuasions, they are critical to the valid testing of all general theory (Shott, 1998). Schiffer's (1988) Fig. 1 showed his view of the role and place of middle-range theory. Figure 1 here is adapted from it, but confined to general (Schiffer's "social") and formation theory. Schiffer's "methodological" theory obviously is important but not in this context. His "inference" theory also is omitted due to uncertainty about its meaning. As above, "correlates" also are important but are distinguished from formation theory. "General" theory is not a general theory but all general theory. Figure 1 shows general and formation theory as independent and equally important domains. Formation theory's dual status requires that it also be situated within general theory by establishing the meaning of archaeological evidence, but in a low position befitting its high empirical content and low comprehensiveness (Schiffer, 1988, Fig. 1) in the context of general theory.
DEFENDING FORMATION THEORY'S IMPORTANCE Even if formation theory is independent, some still question its status. A decade ago, Sabloff et al. (1987, p. 204) described the low regard in which archaeologists held formation theory. If anything, its stature has declined since then. Bradley (1993, p. 131), for instance, feared that formation theory reduces archaeology to technical science. Formation theory, in this view, may be "productive and salutary" (Wobst, 1990, p. 138) at times, but it becomes
Status and Role of Formation Theory
319
an exercise in mere problem solving that distracts archaeologists from the more serious task of theory building at the general level, from the goal of understanding to the narrower one of knowing (Moore and Keene, 1983, p. 4). This sentiment echoes Raab and Goodyear's (1984, p. 262) fear that formation theory would somehow encourage archaeologists to confuse it with general theory. Formation theory is much more than technique, but even technical questions are important in most fields and can advance general theory in any of them, including archaeology. Frankel (1988) objected that formation theory can only identify and therefore reconstruct behavior or process in the short term. Similarly, as substantive uniformitarianism (sensu Gould, 1965), it may remove archaeology's unique rationale by reducing it to the documentation of patterns and processes known already and better through ethnographic observation (Bailey, 1983, p. 173; Moore and Keene, 1983, p. 6). Since archaeology studies the otherwise unknowable past, it registers primarily the nature, rate, and pattern of change in the long term, change that cannot be directly observed in actualistic studies of brief scope. Yet it is arguable to maintain that formation theory somehow is necessary only to the reconstruction of moments in the past. To the extent that it is, then indeed archaeology has small rationale as an distinct discipline. Because behavior and organization are contextual, complex to put it mildly, and perhaps simply arbitrary, Dunnell (1992, pp. 80-81) argued that
Fig. 1. Status and role of formation theory in archaeological practice (adapted from Schiffer, 1988, Fig. 1).
320
Shott
they can be neither described nor explained scientifically. In this respect, Dunnell's view resembles Trigger's, but Dunnell reached different programmatic conclusions. To him science explains only directly observed phenomena, "not some secondary construct" like behavior, which must be inferred. From his selectionist perspective, Dunnell would reject out of hand the validity of formation theory. But to explain the past, Dunnell's selectionism invokes a complex set of behaviors and consequences that are just as secondary and no more directly observable than the cultural organization and behavior that other archaeologists infer (Wylie, 1995, pp. 206-208). Indeed, Dunnell's disciples have argued the need to infer behavior to test their evolutionary theory (O'Brien et al., 1994, p. 260). Eliminating from our purview past behavior in its cultural context would entail a radical reorientation of archaeology that, on present evidence, is not demonstrably superior. What people did in the past in particular, organized, cultural contexts remains a valid archaeological subject, and formation theory is needed to infer it. Criticisms of formation theory are as diverse as contemporary archaeology. Most register legitimate objections to how it is used in service of general theory, especially the limited range of general theory that archaeologists tend to employ. The ontological biases of processual archaeology are undeniable. Indeed, formation theory may be as critical to a more self-conscious engagement of archaeology with modern society as its sociological equivalent is to a comparable engagement there (Mills, 1959). But some question the intrinsic merits of formation theory in the abstract and its critical role in testing general theory. Most such criticism deserves serious regard and some may be valid. None, however, has succeeded in discrediting formation theory as a valid component of archaeological practice.
FORMATION THEORY IN PROFESSIONAL PRACTICE Nevertheless, archaeology has paid lip service more than serious regard to the need to develop formation theory (Sabloff et al., 1987). Binford and Schiffer are exceptions who prove this rule. Prevailing academic practice rewards best work in particular geographic regions, especially those considered glamorous, or particular general theories, often the latest and therefore most fashionable ones, if not the even more fashionable nihilistic repudiation of general theory. Much like flavors in ice cream shops, archaeology programs have their theories du jour. There is nothing wrong intrinsically with such practices except that, among other things, formation theory is thereby marginalized. The discipline's reward structure mirrors
Status and Role of Formation Theory
321
this condition. It is rare for concern with formation theory to be the chief criterion by which grants are awarded or appointments made, let alone interest in a particular aspect of formation theory analogous to the sometimes narrow geographic, theoretical, or technical emphases sought by funders and departments. The denigration of formation theory promotes two related tendencies. The first is to view formation and general theory not just as logically independent but practically incompatible. Archaeologists act on this inclination. We tend to work primarily if not exclusively in one domain or the other, rarely in both at once. The second is to pigeonhole archaeologists as true (i.e., general) theorists or as mere formation-theory technicians (e.g., Bradley, 1993). The tendencies combine in archaeological thought such that formation theory typically is viewed as the province of lithic or faunal analysts and the study of forager societies. In effect, it is considered most relevant to the interpretation of the record produced by societies presumed to be materially determined (e.g., Saitta, 1992, p. 891), and then only by default because there is nothing more of interest there to interpret or understand. Thus Brown's (1995, p. 29) suggestion that lithic analysts, for instance, consider formation theory only because they have little else to consider. Once we turn to more complex cultures and the processes that organized and transformed them, the record somehow can be read as a straightforward document of what happened and why it happened. If it cannot, then at least the assumption is convenient in view of the daunting complexities of assemblage formation. Yet Mehrer (1995) demonstrated the importance of formation theory in validating inference in a complex cultural setting. Hill (1995), Lightfoot (1994), and Mills (1994) did the same for midrange cultures. In each case, formation theory dismissed inviting but misleading interpretations and validated others of substantial anthropological importance. Two factors may explain our collective indifference to formation theory. First is our legitimate interest in cultural—not narrowly archaeological—understanding of the past. Assemblage-formation theory seems an unpleasant distraction from this concern. Second is its apparent refractory quality; we still know far less about the complexities of assemblage formation than we should. If archaeologists grudgingly admit its importance but perceive no way to identify and thus control for it in interpretation, they might not be blamed for neglecting the issue. In his otherwise impressive ceramic analysis, for instance, Blitz (1993, p. 129) made an admirably explicit if unpersuasive admission to this effect. But Lightfoot's (1994) equally impressive research is a good example that showed how formation theory not only may be integrated into ceramic analysis, but that it must be both to extend and to validate cultural inference from the material record. Light-
322
Shott
foot (1994, pp. 81-85) reached conclusions about household size, composition, and cycle—phenomena relevant to many theoretical views—that would have been impossible to achieve without attention to the formation processes that produced the ceramic assemblages he studied. Preoccupation and ignorance explain but do not excuse our indifference, and knowing little about formation theory is poor justification for declining to learn more. Perversely, it is a good way to insure practically against learning more and thereby developing the regard for its importance that formation theory deserves. Thus, some continue to deride formation theory as the archaeological equivalent of Merton's "mere nose-counting" or as a chronicle of the obvious and unimportant. This attitude was captured nicely in Hodder's airy dismissal of formation theory as "processes of deposition and the like" (1986, p. 103)—to be treated shortly in favor of an empathetic contextual approach—and Moore and Keene's (1983, p. 6) talk of "methodological exactitude" in the absence of understanding. Such contrived estrangement of general and formation theory is tragic; it has left archaeology "suspended in space somewhere between animal bones and agit-prop" in Bradley's (1993, p. 131) apt phrasing. The field is ill served by such doctrinaire treatment and the different perceptions of general and formation theory that underlie it. All archaeologists should have an interest in general, that is cultural, theory and its testing against the material record. Yet that record is the product of formation processes (Schiffer, 1987); its reasonable and valid interpretation requires the application of some body of coherent theory that links assemblages in the material record to the cultural behavior and organization that produced them. This much is as true of the material record of complex polities (e.g., Sabloff, 1992, pp. 109-111) as of the lithic and faunal assemblages of early hominids. Formation theory is as relevant to the origins and persistence of complex societies, to the role of human agency in prehistory, and to the mystification of relations of dominance as it is to traditional lithic studies of forager assemblages. No archaeologist can neglect general theory; none can ignore the theoretically grounded inferences attainable only through formation theory that permit us to evaluate that general theory. Widespread indifference to formation theory has predictable consequences. Advocates of opposing theoretical positions talk past one another, not only because they hold different ontological views or because they work narrowly in different areas and so are unaware of others' work, but because none of them, no matter his theoretical persuasion, can demonstrate in the archaeological record the unique validity of his views. Sabloff (1992; see also Sabloff et al., 1987) documented this state of affairs regarding complex so-
Status and Role of Formation Theory
323
cieties in the Maya heartland; Blumenschine et al. (1994) demonstrated the same in a context as distant in time, place, and cultural setting as the study of early hominid behavior. None can do this because none pays nearly sufficient regard to the formation theory that links the existing material record to the cultural past that produced it. Any body of theory can be accommodated to the record depending on how clever we are. But Hodder (1986, p. 96), to the contrary, argument by accommodation resolves nothing, leading only to an unresolved miasma of conflicting theoretical pronouncements. That is a fair description of the current state of the field. This condition is tolerable, perhaps even desirable, to those who are content to advocate by accommodation. But it should be intolerable to any archaeologist who holds reasonable aspirations to recover the past and learn from it constructively rather than merely to indulge his theoretical predilections. Arguments persist indefinitely so long as we cannot distinguish between the alternatives. And we never will so long as we continue to stress multitudes of general theory to the blithe neglect of formation theory.
CONCLUSION Archaeology is not free invention or some literary method, and no serious archaeologist can regard the record as a tabula rasa upon which to inscribe his own version of the past. But absent more serious and systematic attention to formation theory, archaeology continues to drift in the direction that Clarke (1973) legitimately feared. This is not to deny the demonstrated value of humanist, radical, and other orientations. On the contrary, there is no doubt that general theory in archaeology has grown far too narrowly materialist in recent decades or that alternatives are required for a balanced, contextualized understanding of the past. It is merely to emphasize that good archaeology, whatever its point of departure, rests on valid interpretation of the behavior, meaning, and organization that produced the record we study. Because this is so, archaeology requires the permanent recursion of general theory and formation theory.
ACKNOWLEDGMENTS Kent Sandstrom kindly reviewed the manuscript against egregious misrepresentation of the sociological concept of middle-range theory. Eric Henderson directed me to C. Wright Mills' incisive criticism of Parsons. Susan Kent, Michael Schiffer, James Skibo, LuAnn Wandsnider, and Alison Wylie offered criticism that improved the paper. I thank a prominent Brit-
324
Shott
ish archaeologist for suggesting that formation theory is a peculiarly American obsession, not a universal concern. The comment impressed me with the low regard in which formation theory is held. Hill (1995) is among the best recent formation studies, and it is very British. Errors are my responsibility.
REFERENCES CITED Bailey, G. N. (1983). Concepts of time in Quaternary prehistory. Annual Review of
Anthropology 12: 165-192. Bell, J. A. (1994). Reconstructing Prehistory: Scientific Method in Archaeology, Temple University Press, Philadelphia. Bellinger, R. L. (1991). Hunter-Gatherers: Archaeology and Evolution, Plenum Press, New York. Binford, L. R. (1977). General inlroduclion. In Binford, L. (ed.), For Theory Building in Archaeology: Essays on Faunal Remains, Aquatic Resources, Spatial Analysis, and Systemic Modeling, Academic Press, New York, pp. 1-10. Binford, L. R. (1980). Willow smoke and dogs' lails: Hunler-galherer settlement systems and archaeological site formation. American Antiquity 45: 4-20. Binford, L. R. (1981a). Bones: Ancient Men and Modern Myths, Academic Press, New York. Binford, L. R. (1981b). Behavioral archaeology and the "Pompeii premise." Journal of Anthropological Research 37: 195-208. Binford, L. R. (1982). Meaning, inference and Ihe material record. In Renfrew, C, and Shennan, S. (eds.), Ranking, Resource and Exchange, Cambridge University Press, Cambridge, pp. 160-164. Binford, L. R. (1983). Working at Archaeology, Academic Press, Orlando. Binford, S. R., and Binford, L. R. (1968). Archeological theory and method. In Binford, S., and Binford, L. (eds.), New Perspectives in Archeology, Aldine, Chicago, pp. 1-3. Blilz, J. H. (1993). Ancient Chiefdoms of the Tombigbee, Universily of Alabama Press, Tuscaloosa. Blumenschine, R. J., Cavallo, J. A., and Capaldo, S. D. (1994). Compelilion for carcasses and early hominid behavioral ecology: A case sludy and conceptual framework. Journal of Human Evolution 27: 197-213. Bradley, R. (1993). Archaeology: The loss of nerve. In Yoffee, N,, and Sherratt, A. (eds.), Archaeological Theory: Who Sets the Agenda? Cambridge University Press, Cambridge, pp. 131-134. Braun, D. P. (1995). Style, selection, and historicity. In Carr, C., and Neitzel, J. (eds.), Style, Society, and Person: Archaeological and Ethnological Perspectives, Plenum Press, New York, pp. 123-141. Brown, A. G. (1995). Beyond slone-age economics: A slralegy for a contextual lilhic analysis. In Schofield, A. J. (ed.), Lithics in Context Suggestions for the Future Direction of Lithic Studies, Lithic Studies Society, London, pp. 27-36. Clarke, D. (1973) Archaeology: The loss of innocence. Antiquity 47: 6-18. Cowgill, G. L. (1970) Some sampling and reliability problems in archaeology. In Gardin, J.-C.(ed.), ArcMologie et calculateurs: problemes semiologiques et mathematiques, Centre National de la Recherche Scientifique, Paris, pp. 161-175. Cowgill, G. L. (1993) Distinguished lecture in archeology: Beyond criticizing New Archeology. American Anthropologist 95: 551-573. Douglas, M. (1966). Purity and Danger: An Analysis of the Concepts of Pollution and Taboo, Roulledge, London.
Status and Role of Formation Theory
325
Dunnell, R. C. (1992). Is a scientific archaeology possible? In Embree, L. (ed.), Metaarchaeology: Reflections by Archaeologists and Philosophers, Kluwer Academic,
Amsterdam, pp. 75-97. Eggert, M. K. (1982). Comment I: On form and content. In Renfrew, C, Rowlands, M, and Segraves B. (eds.), Theory and Explanation in Archaeology, Academic Press, New York, pp. 139-146. Frankel, D. (1988). Characterizing change in prehistoric sequences: A view from Australia. Archaeology in Oceania 23: 41-48. Fritz, J. M. (1972). Archaeological systems for indirect observation of the past. In Leone, M. (ed.), Contemporary Archaeology: A Guide to Theory and Contributions, Southern Illinois
University Press, Carbondale, pp. 135-157. Goodyear, A. C, Raab, L. M., and Klinger, T. C. (1978). The status of archaeological research design in cultural resource management. American Antiquity 43: 159-173. Gould, S. J. (1965). Is uniformitarianism necessary? American Journal of Science 263: 223-228. Grayson, D. K. (1986). Eoliths, archaeological ambiguity, and the generation of "middle-range" research. In Meltzer, D., Fowler, D., and Sabloff, J. (eds.), American Archaeology Past and Future: A Celebration of the Society for American Archaeology
1935-1985, Smithsonian Institution Press, Washington, DC, pp. 77-133. Hawkes, C. F. (1954). Archeological theory and method: Some suggestions from the Old World. American Anthropologist 56: 155-168. Hill, J. D. (1995). Ritual and Rubbish in the Iron Age of Wessex: A Study on the Formation of a Specific Archaeological Record, BAR British Series, No. 242, Oxford. Hodder, I. (1982). Symbols in Action: Ethnoarchaeological Studies of Material Culture, Cambridge University Press, Cambridge. Hodder, I. (1986). Reading the Past: Current Approaches to Interpretation in Archaeology,
Cambridge University Press, Cambridge, England. Hodder, I. (1991). Postprocessual archaeology and the current debate. In Preucel, R. (ed.), Processual and Postprocessual Archaeologies: Multiple Ways of Knowing the Past, Occasional
Paper, No 10, Center for Archaeological Investigations, Southern Illinois University, Carbondale, pp. 30-41. Ingold, T. (1992). Foraging for data, camping with theories: Hunter-gatherers and nomadic pastoralists in archaeology and anthropology. Antiquity 66: 790-803. Ives, J. W. (1990). A Theory of Northern Athapaskan Prehistory, Westview Press, Boulder, CO. Kelley, J. H., and Hanen, M. P. (1988). Archaeology and the Methodology of Science, University of New Mexico Press, Albuquerque. Kitts, D. B. (1992). The conditions for a nomothetic paleontology. In Nitecki, M., and Nitecki, D. (eds.), History and Evolution, State University of New York Press, Albany, pp. 131-145. Kosso, P. (1991). Method in archaeology: Middle-range theory as hermeneutics. American Antiquity 56: 621-627. Leone, M. P., and Crosby, C. A. (1987). Epilogue: Middle-range theory in historical archaeology. In Spencer-Wood, S. (ed.), Consumer Choice in Historical Archaeology, Plenum Press, New York, pp. 397-410. Lightfoot, R. R. (1994). The Duckfoot Site, Volume 2. Archaeology of the House and Household,
Occasional Paper, No. 4, Crow Canyon Archaeological Center, Cortez, CO. Mehrer, M. W. (1995). Cahokia's Countryside: Household Archaeology, Settlement Patterns, and Social Power, Northern Illinois University Press, DeKalb. Merton, R. K. (1938). Science, Technology and Society in Seventeenth Century England, St.
Catherine Press, Bruges, Belgium. Merton, R. K. (1948). Discussion. American Sociological Review 13: 164-168. Merton, R. K. (1957). Social Theory and Social Structure, 2nd ed., Free Press, Glencoe, IL. Merton, R. K. (1967). On Theoretical Sociology: Five Essays, Old and New, Free Press, New York. Mills, B. J. (1994). Community dynamics and archaeological dynamics: Some considerations of middle-range theory. In Wills, W., and Leonard, R. (eds.), The Ancient Southwestern Community: Models and Methods for the Study of Prehistoric Social Organization, University
of New Mexico Press, Albuquerque, pp. 55-65.
326
Shott
Mills, C. W. (1959). The Sociological Imagination, Oxford University Press, New York. Moore, J. A., and Keene, A. S. (1983). Archaeology and the law of the hammer. In Moore, J., and Keene, A. (eds.), Archaeological Hammers and Theories, Academic Press, New York, pp. 3-13. Nelson, B. A., Kohler, T. A., and Kintigh, K. W. (1994). Demographic alternatives: Consequences for current models of Southwestern prehistory. In Gumerman, G., and Gell-Mann, M. (eds.), Understanding Complexity in the Prehistoric Southwest, Addison-Wesley, Reading, MA, pp. 113-146. O'Brien, M. J., Holland, T. D., Hoard, R. J., and Fox, G. L. (1994). Evolutionary implications of design and performance characteristics of prehistoric pottery. Journal of Archaeological Method and Theory 1: 259-304. O'Connell, J. F. (1995). Ethnoarchaeology needs a general theory of behavior. Journal of Archaeological Research 3: 205-255. Parsons, T. (1948). The position of sociological theory. American Sociological Review 13: 156-164. Peebles, C. S. (1992). Rooting out latent behaviorism in prehistory. In Gardin, J. C., and Peebles, C. (eds.), Representations in Archaeology, Indiana University Press, Bloomington, pp. 357-385. Raab, L. M, and Goodyear, A. C. (1984). Middle-range theory in archaeology: A critical review of origins and applications. American Antiquity 49: 255-268. Sabloff, J. A. (1992). Interpreting the collapse of classic Maya civilization: A case study of changing archaeological perspectives. In Embree, L. (ed.), Metaarchaeology: Reflections by Archaeologists and Philosophers, Kluwer, Dordrecht, The Netherlands, pp. 99-119. Sabloff, J. A., Binford, L. R., and McAnany, P. A.(1987). Understanding the archaeological record. Antiquity 61: 203-209. Saitta, D. J.(1992). Radical archaeology and middle-range methodology. Antiquity 66: 886-897. Salmon, M. (1982). Philosophy and Archaeology, Academic Press, New York. Schiffer, M. B. (1976). Behavioral Archeology, Academic Press, New York. Schiffer, M. B. (1979). A preliminary consideration of behavioral change. In Renfrew, C., and Cooke, K. (eds.), Transformations: Mathematical Approaches to Culture Change, Academic
Press, New York, pp. 353-368. Schiffer, M. B. (1985). Is there a "Pompeii premise" in archaeology? Journal of Anthropological Research 41: 18-41. Schiffer, M. B. (1987). Formation Processes of the Archaeological Record, University of New
Mexico Press, Albuquerque. Schiffer, M. B. (1988). The structure of archaeological theory. American Antiquity 53: 461-485. Shott, M. J. (1993). On recent trends in the anthropology of foragers: Kalahari revisionism and its archaeological implications. Man (n.s.) 27: 843-871. Shott, M. J. (1996a). An exegesis of the curation concept. Journal of Anthropological Research 52: 259-280. Shott, M. J. (1996b). Mortal pots: On use life and vessel size in the formation of ceramic assemblages. American Antiquity 61: 463-482. Shott, M. J. (1998). Crises in forager studies, ethnographic and archeological. Reviews in Anthropology 18 (in press). Stern, N. (1994). The implications of time-averaging for reconstructing the land-use patterns of early tool-using hominids. Journal of Human Evolution 27: 89-105. Taylor, W. W. (1948). A Study of Archeology, Memoir, No. 69, American Anthropological Association, Washington, DC. Thomas, D. H. (1986). Contemporary hunter-gatherer archaeology in America. In Meltzer, D., Fowler, D., and Sabloff, J. (eds.), American Archaeology Past and Future: A Celebration of the Society for American Archaeology 1935-1985, Smithsonian Institution Press,
Washington, DC, pp. 237-276. Trigger, B. G. (1989). A History of Archaeological Thought, Cambridge University Press, Cambridge. Trigger, B. G. (1995). Expanding middle-range theory. Antiquity 69: 449-458.
Status and Role of Formation Theory
327
Tschauner, H. (1996). Middle-range theory, behavioral archaeology, and postempiricist philosophy of science in archaeology. Journal of Archaeological Method and Theory 3: 1-30. Turner, J. H. (1986). The Structure of Sociological Theory, 4th ed., Dorsey Press, Chicago. Varien, M. D., and Potter, J. M. (1997). Unpacking the discard equation: Simulating the accumulation of artifacts in the archaeological record. American Antiquity 62: 194-213. Weber, M. (1958). The Protestant Ethic and the Spirit of Capitalism, Scribner's, New York. Wobst, H. M. (1990). Commentary: A socio-politics of sociopolitics in archaeology. In Pinsky,
V., and Wylie, A. (eds.), Critical Traditions in Contemporary Archaeology, Cambridge University Press, Cambridge, pp. 136-140. Wylie, A. (1989). Matters of fact and matters of interest. In Shennan, S. (ed.), Archaeological Approaches to Cultural Identity, Unwin Hyman, London, pp. 94-109. Wylie, A. (1995). An expanded behavioral archaeology: Transformation and redefinition. In Skibo, J., Walker, W., and Nielsen, A. (eds.), Expanding Archaeology, University of Utah Press, Salt Lake City, pp. 198-209. Yellen, J. E. (1989). The present and the future of hunter-gatherer studies. In Lamberg-Karlovsky, C. (ed.), Archaeological Thought in America, Cambridge University Press, Cambridge, pp. 103-116.
BIBLIOGRAPHY OF RECENT LITERATURE Aldenderfer, M. (1977). The Computer Simulation of Assemblage Formation Processes: The Evaluation of Multivariate Statistical Methods in Archaeological Research, Unpublished Ph.D. dissertation, Department of Anthropology, Pennsylvania State University, State College. Ammerman, A. J., and Feldman, M. (1974). On the "making" of an assemblage of stone
tools. American Antiquity 39: 610-616. Baxter, M. J., and Cool, H. E. (1994). Notes on Some Statistical Aspects of Pottery Quantification, Research Report 15/94, Department of Mathematics, Statistics, and Operational Research, Nottingham Trent University. Bettinger, R. L. (1987). Archaeological approaches to hunter-gatherers. Annual Review of
Anthropology 16: 121-142. Binford, L. R. (1983). In Pursuit of the Past, Thames and Hudson, London. Byrd, J. E., and Owens, D. D. (1997). A method for measuring relative abundance of fragmented archaeological ceramics. Journal of Field Archaeology 24: 315-320. Cameron, C. M., and Tomka, S. A. (eds.) (1993). The Abandonment of Settlements and
Regions: Ethnoarchaeological and Archaeological Approaches, Cambridge University Press, Cambridge. Childress, M. (1992). Mortuary vessels and comparative ceramic analysis: An example from the Chucalissa site. Southeastern Archaeology 11: 31-50. Cool, H. E., and Baxter, M. J. (1996). Quantifying Glass Assemblages, Research Report 5/96, Department of Mathematics, Statistics, and Operational Research, Nottingham Trent University. Cordell, L. S., and Upham, S. (1989) Culture and cultural behavior: One more time, please. American Antiquity 54: 815-819. Cordell, L. S., Upham, S,, and Brock, S. L. (1987). Obscuring cultural patterns in the archaeological record: A discussion from Southwestern archaeology. American Antiquity 52: 565-577. Cowgill, G. L. (1989). The concept of diversity in archaeological theory. In Leonard, R., and Jones, G. (eds.), Quantifying Diversity in Archaeology, Cambridge University Press, Cambridge, pp. 131-141. DeBoer, W. R. (1983). The archaeological record as preserved death assemblage. In Moore, J., and Keene, A. (eds.), Archaeological Hammers and Theories, Academic Press, New York, pp. 19-36.
328
Shott
DeBoer, W. R., and Lathrap, D. (1979). The making and breaking of Shipibo-Conibo ceramics. In Kramer, C. (ed.), Ethnoarchaeology: Implications of Ethnography for Archaeology, Columbia University Press, New York, pp. 102-138. Depew, D. J., and Weber, B. H. (1995). Darwinisn Evolving: Systems Dynamics and the Genealogy of Natural Selection, MIT Press, Cambridge, MA. Dibble, H., and Holland, N. (1992). On assemblage variability in the middle paleolithic of western Europe: History, perspectives and a new synthesis. In Dibble, H., and Mellars, P. (eds.), The Middle Paleolithic: Adaptation, Behavior and Variability, University Museum Press, Philadelphia, pp. 1-28. Dunnell, R. C. (1992). The notion site. In Rossignol, J., and Wandsnider, L. (eds.), Space, Time, and Archaeological Landscapes, Plenum Press, New York, pp. 21-41. Dunnell, R. C. (1992). Archaeology and evolutionary science. In Wandsnider, L. (ed.), Quandaries and Quests: Visions of Archaeology's Future, Southern Illinois University Press, Carbondale, pp. 209-224. Ebert, J. I. (1992). Distributional Archaeology, University of New Mexico Press, Albuquerque. Embree, L. (1987). Archaeology: The most basic science of all. Antiquity 61: 75-78. Gamble, C. (1986). The Palaeolithic Settlement of Europe, Cambridge University Press, Cambridge. Goldberg, P., Nash, D. T., and Petraglia, M. D. (1993). Formation Processes in Archaeological
Context, Prehistory Press, Madison, WI. Hayden, B., and Cannon, A. (1983). Where the garbage goes: Refuse disposal in the Maya highlands. Journal of Anthropological Archaeology 2: 117-163. Keegan, W. F. (1991). Culture processes and culture realities. In Preucel, R. (ed.), Processual and Postprocessual Archaeologies: Multiple Ways of Knowing the Past, Occasional Paper, No. 10, Center for Archaeological Investigations, Southern Illinois University, Carbondale, pp. 183-196. Kidwell, S. M., and Bosence, D. W. (1991). Taphonomy and time-averaging of marine shelly faunas. In Allison, P., and Briggs, D. (eds.), Taphonomy: Releasing the Data Locked in the Fossil Record, Plenum Press, New York, pp. 115-209. Kristiansen, K, (1985). The place of chronological studies in archaeology: A view from the Old World. Oxford Journal of Archaeology 4: 251-266. Kristiansen, K. (ed.) (1985). Archaeological Formation Processes: The Representativity of Archaeological Remains from Danish Prehistory, Nationalmuseet, Copenhagen. Lockyear, K. (1993). Coin hoard formation revisited. In Andresen, J., Madsen, T., and Scollar, I. (eds.), Computing the Past: Computer Applications and Quantitative Methods in Archaeology, Aarhus University Press, Aarhus, Denmark, pp. 367-376. Mills, B. J. (1989). Integrating functional analysis of vessels and sherds through models of ceramic assemblage formation. World Archaeology 221: 133-147. Needham, S., and Spence, T. (1997). Refuse and the formation of middens. Antiquity 71: 77-90. Oetelaar, G. (1993). Identifying site structure in the archaeological record: An Illinois Mississippian example. American Antiquity 58: 662-687. Orton, C. R. (1993). How many pots make five? An historical review of pottery quantification. Archaeometry 35: 169-184. Pauketat, T. R. (1989). Monitoring Mississippian homestead occupation span and economy using ceramic refuse. American Antiquity 54: 288-310. Pollard, J. (1995). Inscribing space: Formal deposition at the later Neolithic monument of Woodhenge, Wiltshire. Proceedings of the Prehistoric Society 61: 137-156. Potter, P. B. (1992). Middle-range theory, ceramics, and capitalism in 19th-century Rockbridge County, Virginia. In Little, B. (ed.), Text-Aided Archaeology, CRC Press, Boca Raton, pp. 9-23. Pyszczyk, H. (1984). Site occupation span as a factor in artifact assemblage variability and frequency. In Burley, D. (ed.), Archaeology in Alberta 1983, Occasional Paper, No. 23, Archaeological Survey of Alberta, Edmonton, pp. 60-76.
Status and Role of Formation Theory
329
Reid, J., Schiffer, N., Whittlesey, S., Hinkes, M., Sullivan, A., Downum, C, Longacre, W., and Tuggle, H. (1989). Perception and interpretation in contemporary Southwestern archaeology: Comments on Cordell, Upham, and Brock. American Antiquity 54: 802-814. Schlanger, S. H. (1990). Artifact assemblage composition and site occupation Duration. In Minnis, P., and Redman, C. (eds.), Perspectives on Southwestern Prehistory, Westview, Boulder, CO, pp. 103-121. Shott, M. J. (1989). Diversity, organization, and behavior in the material record: Ethnographic and archaeological examples. Current Anthropology 30: 283-301. Shott, M. J. (1989). On tool class use lives and the formation of archaeological assemblages. American Antiquity 54: 9-30. Skibo, J. M., Walker, W. H., and Nielsen, A. E. (eds.) (1995). Expanding Archaeology, University of Utah Press, Salt Lake City. Steponaitis, V. P., and Kintigh, K. W. (1993). Estimating site occupation spans from dated artiface types: Some new approaches. In Stoltman, J. (ed.), Archaeology of Eastern North America: Papers in Honor of Stephen Williams, Archaeological Report, No. 25, Mississippi Department of Archives and History, pp. 349-361. Stern, N. (1993). The structure of the Lower Pleistocene archaeological record: A case study from the Koobi Fora Formation. Current Anthropology 34: 201-225. Stevenson, M. G. (1982). Toward an understanding of site abandonment behavior: Evidence from historic mining camps in the southwest Yukon. Journal of Anthropological Archaeology 1: 237-265. Varien, M. D., and Mills, B. J. (1997). Accumulations research: Problems and prospects for estimating site occupation span. Journal of Archaeological Method and Theory 4: 141-191. Villa, P. (1983). Terra Amata and the Middle Pleistocene Archaeological Record of Southern France, University of California Press, Berkeley. Wandsnider, L (1989). Long-Term Land Use, Formation Processes, and the Structure of the Archaeological Landscape: A Case Study from Southwestern Wyoming, Unpublished Ph.D. dissertation, Department of Anthropology, University of New Mexico, Albuquerque. Watson, R. A. (1990). Ozymandias, king of kings: Postprocessual radical archaeology as critique. American Antiquity 55: 673-698. Western, D. (1980). Linking the ecology of past and present mammal communities. In Behrensmeyer, A., and Hill, A. (eds.), Fossils in the Making: Vertebrate Taphonomy and Paleoecology, University of Chicago Press, Chicago, pp. 41-54. Willey, G. R., and Sabloff, J. A. (1980). A History of American Archaeology, 2nd ed., Freeman, San Francisco.