find myself, at the outset of this paper, in the uncomfortable position of having to preface this presentation with two apologies and an explanation. This unfortunate turn of events results from the fact that, in preparing this paper, I have violated my own cardinal rules of paper presentation. The first of these rules states, in no uncertain terms, that one must never, but never, use the word epistemology in a paper title. Hence I come to my first mea culpa; in my own defense I can only offer the feeble, and perhaps partially exculpatory, observation that the offending word does, at least, appear after the colon. The second transgression, serves, I fear, only to deepen the effect of the first. No slides. In fact no visual aids of any kind! A paper that openly advertises a discussion of epistemology and offers no visual diversions to the wandering mind--Oh my! I feel it my duty, as partial atonement, to pause here until the rush for the exit subsides, before proceeding to the promised explanation.

Why offer a discussion of epistemology, a topic which is, after all, only slightly less distasteful to all sane historical archaeologists than a discourse on the dreaded whiteware/pearlware controversy? I can only suggest that the trajectory of a rapidly maturing field, such as ours, increasingly necessitates a reflexive attitude on the part of it's practices. Specifically, resent provocative contributions in the field focus attention on the need to develop productive theoretical positions to guide historical archeology in the future.

In the last five years our field has witnessed an increased interest in theoretical approaches which can be characterized as Marxist or Marxist influenced. This trend is evidenced both by papers presented at recent professional meetings and by published contributions. If we examine this body of work we find approaches which can be identified as explicitly Marxist, as well as those which acknowledge, to one degree or another, a Marxist paternity, while eschewing a strict reading of Marx. Some of these derivative approaches aim at correcting what are perceived to be defects in Marx's approach, others proceed by emphasizing one or another element drawn from Marx. The focus here will be on these derivatives---on what might be termed the partitive application of Marxist theory.

Two approaches drawn from this rapidly growing body of work will be discussed here--that of Charles Orser and a recent contribution by Mark Leone and his colleges at Historic Annapolis. These examples, I think, represent particularly original and stimulating examples of the two trends in derivative Marxism outlined above. These examples will be examined in the light of Marx's epistemology. However, by highlighting broad questions of theoretical underpinnings the attention to this work will, I think, be of some interest even to those who have no sympathy with the specific approaches discussed.

Orser finds historical materialism an approach "...well suited to archaeological research because of the nature of the information collected" (1987:122). However the `political' element in Marx remains, for him, unsettling. While acknowledging the central role that Marx and his followers accord to political action, indeed to revolution, Orser concludes that " scholars, historical archaeologists would gain nothing by tying themselves to a socialist program...." (1987:123).

Instead what Orser wishes to construct is a non-Marxist historical materialism which aims to "...use and improve [Marx's] good ideas and discard the rest" (1987:123). The result of this program of revision is to produce, in Orser's words, an "...historical materialism [which] rests on research and scholarship rather than on dogma and rhetoric" (1987:123).

Now, this is a provocative position. It challenges us to examine the place of politics in Marx. Is political action mere `dogma and rhetoric' external to historical materialism? Can we identify a `political Marx' and a `scholarly Marx'? What are the implications of de-politicizing Marxism?

Inscribed on Marx's headstone in Highgate Cemetery are two short excerpts from the voluminous output of his busy life. I am sure most of you immediately guessed one of the two. It is in fact that famous quotation from the Communist Manifesto [1848], perhaps the most famous appeal to revolution in human history. While this ringing clarion call is testimony to the importance of political practice in Marx's theory, the second inscription on the headstone is of more interest to us in the present context. The last of eleven of the Theses on Feuerbach, this single sentence remains key to understanding Marx. It states simply: "The philosophers have only interpreted the world, in various ways; the point is to change it" (Marx 1947:123).

More than just a rhetorical assertion of a political dogma this passage cuts to the heart of Marx's historical materialism and lays bear the epistemological underpinning that lie at the heart of his life's work. The second of the theses should help clarify this point:

"The question whether objective truth can be attributed to human thinking is not a question of theory but is a practical question. Man must prove the truth, i.e. the reality and power, the this-sidedness of his thinking in practice. The dispute over the reality or non-reality of thinking that is isolated from practice is a purely scholastic question" (Marx 1947:121).

Now, in order to understand this passage, and indeed all of Marx's work after 1845, we must remember that Marx's project was to construct a scientific socialism. But Marx's science was a non-positivist materialism; it was a science which he saw as absolutely anchored in a unity of theory and practice (cf. Diamond 1975:871). The two stand as a dialectic, and as such have no meaning when considered separately. The `proof' of theory lies in practice--in the real world.

Marx detailed his disaffection with Comtean positivism in his re-telling of the Roman myth of Cacus. Fathered by Vulcan, Cacus was half human, half demon---with a taste for livestock. Our story concerns Cacus' ingenious means of sating this appetite. The beast emerged from his cave for nocturnal foraging expeditions, and to evade detection, induced his pray to walk backward into his lair. The puzzled owners of the missing oxen traced the foot prints and could only conclude that their oxen had emerged from the cave and, after walking to the middle of the field, had simply disappeared into thin air!

As Bertell Ollman has pointed out:

If the searchers had studied methodology in an American university they might have counted the footprints and measured the depth of each step, but they would have arrived at the same conclusion. Focusing exclusively on appearances, no matter how carefully, concentrating entirely on the evidence which strikes us directly, can be extremely misleading. How typical is the error found in this example? According to Marx, this is how most people in our society understand the world. Basing themselves on what they see, hear, and bump into in their immediate surroundings, they often arrive at conclusions which are the exact opposite of the truth (1986:42-43).

The nineteenth century positivist tradition derived from Comte and Hume attempted to control for human subjectivity by discounting all but the directly perceived characteristics of things. Thus allowing researchers to reach a consensus on the truth value of their research. According to this approach there can be, at least in theory, no assumptions about the reality of non-sensory phenomenon behind these perceptions, or even causal links between co-occurring observations. Now, clearly, a strict adherence to this approach precludes the consideration of such non-perceptible phenomenon as ideology. In practice, as far as social science goes, our scientific project is reduced to a through going behaviorism. Human agency and causality must, then, lie in the twilight realm of the unknowable---the speculative.

On the other hand at the core of Marx's method is the dialectic. There is perhaps more confusion and misunderstanding surrounding this aspect of Marx than any other element of his work. Indeed, it is a problematical concept, and yet it is one which must be grappled with in order to begin to understand Marx. It is perhaps best to begin by describing what the dialectical method is not. Here I again turn to Bertell Ollman:

Dialectics is not a rock-ribbed triad of thesis-antithesis-synthesis which serves as an all purpose explanation, nor does it provide a formula that enables us to praise or predict anything, nor is it the motor force of history. The dialectic as such explains nothing, proves nothing, predicts nothing, and causes nothing to happen. Rather, dialectics is a way of thinking which brings into focus the full range of changes and interactions that occur in the world (1986:42).

Dialectics encourages us to think in ways that are antithetical to the `common sense' notions we hold as heirs to a positivist tradition. We are encourages to think about how we conceptualize the world and about how we cut up the world into discrete packages for contemplation and study. A phenomenon (capital or wage labor, for instance) is not viewed as a isolated entity with a history and with connections to other discrete entities. Rather, capital or wage labor are processes which contain their histories and futures and their relations with other parts of the whole. On the one hand, and here Marx reveals part of his debt to Hegel, being is becoming and, on the other, reality is not packaged in neat, easily digested units, suitable for consumption by intellectually dyspeptic academics.

This way of thinking has numerous advantages. Unlike the simple functionalism that underlies much of positivist science we need not constantly search for external `kickers' to animate our static constructs. Change is not an extraneous and essential random intrusion from outside the system. Dialectics does not, necessarily, discount random events. However, when such events do occur they can be used to illuminate the internal structure and relations of the phenomenon under study.

By focusing on the processes and relationships behind the little packages of reality we are concerned with it is possible to see beyond often misleading surface appearances. The result is that the `commons sense' notions we hold as participants in a given culture are frequently exposed as illusory. For instance, capitalists (and classical economists) may view rent, profit and interest as they appear to the capitalist, as distinct and natural phenomena, and not see their central unity in a historical, and transitory, mode of production. The point of science must be to uncover the meaning hidden beneath the surface appearance of things; as Marx has pointed out `science would be superfluous if the outward appearance and the essence of things directly coincided' (quoted in Wolf 1981:44).

Striped of the political dimension and the dialectic Marx's formulation looses its epistemological grounding, the researcher then is responsible for filling this void. All too often without these two key elements we are left with a mechanical materialism tied to a positivist methodology. To return to Charles Orser's work we must note that he is careful to avoid this obvious trap, and, consequently, sidesteps the vulgar materialism of Marvin Harris, he acknowledges the complexity of the dialectical relationship between thought and action. However, by dismissing the central dialectic in Marx, that is the unity of thought and action, he removes the underpinnings of historical materialism and offers no replacement, leaving the epistemological foundations of his work unexamined.

In a recent article Leone, Potter and Shackel have advocate an approach to historical archeology they term `critical archeology'. The authors tip their hats to the Frankfurt school, a philosophical tradition in sociology, which extracted from Marx the notion that knowledge, as an ideological production, is always linked to a given historical moment. As such knowledge must be subject to critical examination in the light of the social forces that produce it. For Leone and his colleagues this means that: "Critical theory asks of any set of conclusions from what point of view they are constructed" (Leone, Potter and Shackel 1987: 284). The authors then go on to note: "The question is intended to help establish their degree of validity" (1987:284).

It is, however, precisely on this second point, the question of validation, that problems arise. How are we to evaluate conflicting interpretations of the past? Unless we have a clear notion of what validation consists of, of how we can sort through these various accounts of the past, critical archeology will only succeed in obscuring the world in a subjectivist haze and we will find that `all that is solid melts into air'.

Leone and his colleagues are aware of this dilemma. After acknowledging that "[p]ositivism has come under criticism for frequently producing knowledge so narrow as to seem irrelevant.." they then go on to suggest that "positivism is be abandoned but rather to be adjusted to the realities of archaeological data".

What such an `adjustment' would entail is not specified and the authors note that "...positivism and critical theory in archaeology have not yet been synthesized..."(285). The presentation and analysis of archaeological data that follows this observation fits very well into an unsynthesized positivist tradition. The authors "...argue that segmentation and standardization in ceramic use in Annapolis accompany an etiquette associated with the accumulation of profit" and that "[b]oth segmentation and standardization will be reflected in the material culture used in many aspects of daily life, including the table and its setting, as people are incorporated into the life of capitalism" (288).

In order to test this hypothesis a formula is constructed to express, as a numerical value, the variability in ceramic vessel types and sizes within archaeological assemblages. This formula is then applied to assemblages from three sites and various time periods which represent various "socioeconomic characteristics". The results of this test lead the authors to conclude that "[t]he greater variety of dish sizes and wares in the archaeological record reflects a new etiquette, an increasing segmentation at the table that served both as a training ground for the new order and as reinforcement for it" (289).

Now, I do not suggest that these results are `wrong', indeed, I believe them to be potentially quite useful. However, on there own, as presented by their authors, they only serve to illustrate the limitations of the positivist approach. The demonstration of this ceramic variability tells us nothing about the meaning of this evidence of segmentation. It in fact displays the classic pit falls of positivism discussed above. A co-variation between two phenomena is displayed (in this case time and variability in a class of material culture) but the cause or meaning of this co-variation is not in any way verified or explained.

We are left with an unexamined functionalist tautology: segmentation is a prerequisite of capitalism, and is simultaneously reason enough for its appearance. In addition we are led to a reactive and normative view of the human actors in the historical drama. We are left to conclude that people are simply passive receptacles for an ideology (segmentation) which is massively over determined by the economic base (capitalism).

When we look at the results of this work we find that, on the one hand, a positivist approach in no way validates a given view of history--the meaning the authors assign to the evidence for segmentation is, in fact, external to the hypothesis tested. One could substitute any number of alternate explanations for the observed co-occurrence, because, in terms of the positivist paradigm, no causal link is being tested. On the other hand, ironically, the result of this exercise is a sterile behaviorism.

The irony lies in the fact that Leone and his colleagues suggest that the approach they propose allows for the study of consciousness; that it can help strip away ideological constructions of history that act to mask conflict and contradiction. In fact the authors are in danger of erecting their own construction which serves to obscure what they are attempting to expose. We are offered no notion of the class specific experience of segmentation, nor of any tensions, contradictions or resistance to the attempted imposition of the new `etiquette' of segmentation.

Let me emphasize, once again, that the problem is not with the empirical evidence Leone and his colleagues offer---rather the problem lies with the way these data are approached.

For Marx the research imperative lay in "ascending from the abstract to the concrete" (Wolf 1981:44). As Eric Wolf explains it:

The road from the abstract formulation to the explication of the concrete leads over a series of conceptual approximations to the explication of particular phenomena occurring in particular places and times. Marx clearly characterized the first two steps in approximation. The concept of the labor process, and the concept of the mode of production. A third step in approximation is to conceptualize a social formation, a historically concrete `society' embodying one or more modes of production.... A fourth step or level of approximation is the observation and interpretation of social interaction and the cultural forms that mediate it....(1981:44-45)

If we cut short this process and, instead, proceed directly to the concrete, to empirical observations, we run the risk of being mislead by the surface appearances of things. It becomes almost inevitable that we will fall victim to a distorted understanding molded by our unexamined cultural baggage.

It is necessary, I fear, to return once more to the question of verification. If positivist methodology can only nibble away at the margins---if it is limited to assembling a series of co-occurrences, correlations and systematic relationships between phenomenon---is our only alternative abject surrender to the hobgoblin of subjectivist nihilism?

To avoid this dead end we must begin by discarding the notion, inherent in the positivist approach, that truth can be arrived at by the simple application of a formal set of rules to little bits of the world. Marx, instead, attempted to construct a totalizing system of thought that explained the underlying principles that structure social reality. Verification of this system, in his view, could come only through praxis---the dialectical interaction between thought and action.

As Bridget O'Laughlin puts it: "To some extent then Marxist anthropology must be applied anthropology, the university and the classroom a locus of political struggle, and praxis an essential aspect of verification" (1975:369). Thus, for Marx, politics and academic theory can never be separated; the one tests the validity of the other. Only in the dialectical unity of the two is there escape from the Scylla and Charybdis of empty empiricism, on the one hand, and bland subjectivism, on the other.

In conclusion, let me hasten to add, least this paper be taken as a plea for orthodoxy, that such is not my intention. Contrary to popular prejudice Marxism is not a closed, dogmatic, system. Marx, himself, never considered his theoretical work complete, or unalterable. Indeed, when, even in his own lifetime, he detected the dead hand of orthodoxy descending on his work he declared he was not a Marxist. What is most remarkable in his legacy is the tension between, on the one hand, a totalizing system and, on the other, the requirement that that body of work be responsive to the lessons learned through engagement with a real and inevitably changing world.

The challenge that I take from my reading of Marx is that we all, Marxist and non-Marxist alike, must in and through our work confront a series of vital questions: What epistemological assumptions anchor our theory? What are the real world implications of what we say and do?