A site that supports both Classless Society and Direct Democracy
One cannot exist without the other
Ugug and Dudu invent class conflict:
the significance of classless societies for Marxist theories
of class struggle and transition [*]
by Larry M Miller
"An explanation should be as simple as possible, and no simpler."
most obvious characteristic of Marxism, now, is the multiplicity of Marxisms,
neo-Marxisms and post-Marxisms that litter the landscape. It has been a long
time since we could talk of “Marxism” as a unified theory or one, which
could compel substantial agreement about even fundamental concepts and
priorities. There is no single argument whose resolution will resolve the
disagreements that divide us. That kind of unity is a pointless hope and in
all likelihood, undesirable. Still, the infinite multiplication of points of
disagreement is equally undesirable. We must continue to develop criteria to
clarify theoretical disagreements, by which we can judge among alternative
concepts, not so that we can forge some unwanted neo-orthodoxy, but so that we
can focus dispute and research where it will be most fruitful. While it is
likely that at this stage of our intellectual history and for some time to
come, more than one position will be “right”, it is even more likely that
many positions are wrong. So, we need criteria for choice.
of the reasons why we have such a multiplication of possible theories is that
we are trying to fit our theories to too narrow and limited a range of
historical forms. We can cut down on the range of theoretically coherent
options even if we cannot resolve all debates, if we can demonstrate that some
Marxisms produce concepts that explain a more substantial field of history
than others. Only by extending the field to include the entire history of
class society, are we defining the field in a way that is fundamentally
compatible with the theoretical and political tasks set for the theory at the
beginning, by Marx and Engels, the vision of the purpose of our work which may
be the one thing which we do, indeed, all share. We will find that working
with such an extended field generates concepts that help us to choose among
theories, leaving us with a richer, fuller, more elegant theory of
contemporary societies and one which is more obviously applicable to
theorizing about any transitions to socialism in the future. By defining the
field as the entire history of class society we include its origins and
potential disappearance. We therefore need definitions of class, theories of
class structure and of class struggle, which do not merely account for the
behavior of social groups within capitalism, or in the context of the
transition from feudalism to capitalism, or any other sort of transition from
one sort of class society to another, but which can account for the origins
and trajectory of class society in general.
of the weaknesses of many contemporary Marxisms is that they tend to define
class and class struggle in ways which are much too directly tied to the
dynamics of the mode of production that the theory is most interested in
explaining. By defining the task partially, the work only gets partially done.
Many of the arguments between the competing Marxisms have to do with the
exhibition of one or another partially correct or partially useful insight in
the absence of problems that might have posed the significance or usefulness
of something that another partial definition of the task would have revealed.
By defining the task as broadly as possible, we set up criteria for the
evaluation of theory that test it against the widest range of historical
situations, and give, by that standard alone, criteria for sorting amongst
some of the contemporary arguments.
point of this paper is to explore the significance of theorizing classless
societies and the transition from classless to class forms of society for the
development of Marxist theory. Some of the conceptual shortcomings of recent
theoretical work are rooted in this problem of too narrowly contemporary a
focus. One example, which comes readily to mind, is the debate among Marxist
theorists of the relative significance of economic issues,
"exploitation", as opposed to workplace issues, "control",
in generating class contradictions, etc. Over the last few years this argument
has been explored from several points of view. I have no intention of
exploring it in detail here, but merely calling attention to some notable and
varied examples will demonstrate my point. Take for example, Roemer's argument
that "property relations are the key to exploitation, not
social relations at the point of production"
and that "democratic control of the surplus, not of the
workplace is the real necessity for social transformation."
Roemer attempts to escape from the limits of talking only about
capitalism. He "wanted to abstract from actually existing capitalism as
much as possible to construct a general theory that could be used to discuss
exploitation under socialism as well."
But, Roemer's argument proceeds logically, not sociologically by
constructing extreme cases. He is not totally off the mark. Of course,
exploitation is more central than the organization of work in constituting the
nature of capitalism. But, Roemer escapes only from the historical
specificities of particular capitalisms, not from capitalist categories in
general. Property relations are a central constituting fact of capitalist
class society, but not of every class society (e.g. Sumer). Roemer is an
example of the awkwardness that gets introduced into theorizing about the
transition from the capitalist mode of production to non-class societies
("socialism", "communism") when the focus of the
theorizing is only on capitalist society itself. You cannot theorize socialism
using only capitalist categories.
and Wolff share Roemer's insistence that social relations at the point of
production are not what generate class in contemporary society, although they
focus on the economics of surplus extraction and not on property relations per
se. In contrast, Michael Buroway emphasizes the sociology of the point of
The organization of work, for him, generates "the micro
foundations of class".
Buroway selfconsciously counterpoises his work to another recent
strain in Marxist writing, the critique of Marxism as a political project from
the point of view of democracy. This too, takes many, mutually incompatible
but the common theme is that political democracy and economic
equality are both goods, but that Marxism has over the years sacrificed a
concern for the former in the interest of the latter. In its most extreme
form, e.g. Laclau and Mouffe or Bowles and Gintis, democracy is, in one or
another way opposed to economic equality as a more historically privileged
area of struggle or the wave of the future etc. The political impulse of a lot
of this arguing is evident. But how can it be evaluated as theory. It is the
contention of this paper that by trying to think through these questions from
the point of view of the longest duree of all, the entire history of class
society, it will indeed be possible to do what Roemer attempted and failed at,
produce a general theory adequate to the task of thinking about the transition
from class to classless society and criticizing the existing (and no-longer
existing) forms of socialism as well as capitalism from that perspective.
cannot adequately conceptualize the transition from class to classless society
on the basis of categories overly rooted in capitalist society or, at best, in
the transition from feudalism to capitalism. Excessive focus on one epoch,
even our own, helps prevent the achievement of conceptual clarity in a project
which is nothing if it is not a general theory of history. Categories that are
too historically specific are inadequate to the task of a historically
comparative general theory. While it is important to reject trans historical
categories that are not rooted in any concept of historical experience, it is
equally necessary to generate trans epochal concepts that can be used to
explain a process of transition or compare one epoch or social formation with
another. It is precisely the attempt to analyse, for example, soviet socialism
with either a historical categories or categories derived from the study of
capitalism alone (or at best the transition from feudalism to capitalism)
which impoverishes Marxist theorizing about the transition to socialism.
focusing on class society it is possible to come to the conclusion that the
form of exploitation is the key to understanding the form of the society. But
this does not makes sense heuristically if one is addressing the analysis of
classless society. One could, I suppose, simply lump all classless societies
together, but this evades significant questions. To develop concepts that will
cover both classless and class societies, it is necessary to recognize that
the organization of surplus extraction, central as it is to an understanding
of class societies has to be a conceptual subset of something else which will
allow the analysis of different forms of social organization in which surplus
extraction does not occur. How else can one differentiate between the numerous
different types of societies in which in Resnick and Wolfe's terms
"surplus is consumed by those who produce it"
or in Polanyi's terms "societies in which surplus is not
produced at all"
(not exactly the same sets). Is there any point in declaring that
for Marxist theory there is no conceptual difference between the hunting and
gathering societies of the Central African rain forest, the horticultural
societies of Melanesia and the almost, but not quite, class divided societies
of early Sumer? Clearly, one thing we do need is a conceptual category to
recognize their similarities. For Marxist theory that concept is
classlessness. Equally clearly, the differences between these societies are
pronounced and figure significantly in what we would expect their future
development to be. It does not make sense to insist, as Resnick and Wolff
would appear to require, that the distinctions among these various societies
be seen as restricted to the political, cultural or ideological. Undoubtedly
some of the differences are best analysed in those terms but many of the most
salient distinctions in clarifying the significance of these societies
analytically are clustered around a set of issues that have been privileged in
Marxian thought since the drafting of The German Ideology, namely the
organization of production and reproduction -especially in the area of the
organization of work. Is work organized and performed individually or
collectively, by the household or by the group? Is consumption restricted
/contained within the immediate producing unit or do groups of producers
create surplus which is then redistributed among them according to some set of
rules which, nevertheless retains the surplus more generally
within the larger
group of those who produce it? The general sense then, is that in classless
societies it is the organization of work which differentiates modes of
production- in class societies it is the organization of surplus extraction.
In classless societies that appear to produce a surplus, e.g. the "big
man" type of societies of Melanesia, the organization of surplus movement
is subordinated to the organization of work. The shift of predominance from
work to surplus extraction as the organizing set of issues is what
characterized the transformation of classless into class societies. But that
means that we need to look at the organization of both work and surplus
transfer (where it exists) in both cases and not privilege one and ignore the
Clearly, in order to discuss the transition from classless society to any form of class society we need definitions of class which will work in pre industrial and non capitalist contexts. The concept of class cannot be developed alone or in isolation. Class is a relational category. Thus the concept of class depends on the concept of `mode of production’, which specifies the nature of the relationship. So what do we mean by mode of production? First of all, it is helpful to specify what kind of a concept mode of production is. It is, of course, an abstraction. One does not look for modes of production to appear purely and simply in history. Actual modes of production can be understood as the limits of variation of particular social relations of production. Primarily an analytic construct, it is also a historical category.  Thus, the mode of production is not a Weberian `ideal type'. But when we extract particular modes of production from historical circumstance and talk about them individually as if the only determinations that mattered were those social relations that constituted the particular mode of production then in a sense we are talking about ideal types of modes of production in the Weberian sense. We are looking at social relations in a vacuum. The ideal type of the mode of production is useful as a tool of analysis to help us recognize what is going on- but it is not the thing itself. The mode of production is a complex structured element of a total society/social formation, a structure of social relations.
OF WORK, SURPLUS, REPRODUCTION
sets of social relations are particularly important in recognizing and
defining particular modes of production. Since the Critique of Political
Economy, the traditional terms among Marxists for discussing this have been
relations of production and forces of production. Since Cohen
this language and the technological determinism it fosters have
had a strong revival in some quarters. This has had the unfortunate tendency
to focus the search for contradiction between the social and the material
whereas the real issue is in contradiction between structures of social
relations. Thus I prefer to talk (without claiming any particular novelty of
content) of a mode of production as composed of the social organization of
surplus appropriation, the social organization of work, and the social
organization of reproduction.
relations constitute, in the most abstract sense, a particular mode of
production, but they do not function as equals. In any given mode of
production, one will be determinate. This does not mean that influence will
move only in one direction, but that the degree to which either the extraction
of surplus or the social organization of work and the social organization of
reproduction will affect each other and all other social relations will be
are two fundamental sets of relationships, which constitute any mode of
production. One is the organization of surplus appropriation. The other is
composed of the organization of work and the organization of reproduction.
These latter two exist in every mode of production, class and classless and
are therefore, in a certain sense primary. Nothing exists without them. Work
is done, and therefore, somehow, organized, in any society. If the next
generation is not reproduced, the society ceases to exist. That, too, must be
organized in some fashion. There are, though, some societies, and therefore
some modes of production, in which no surplus is produced, e.g. subsistence
level foraging people like the Mbuti in central Africa. Some work in an Mbuti
band is organized around the individual household, e.g. gathering, food
preparation and consumption, and the care of children who cannot yet walk
(below the age of about one). Some work is organized collectively, e.g.
hunting and the rearing and educating of children who can walk (above the age
of about one).
could say that there are two modes of production present in an Mbuti band: a
household mode and a collective mode, with the collective mode being dominant.
This contrasts with the situation in some other foraging societies, e.g. the!
Kung or Ju/Wasi, where there appears to be very little in the way of
collective organization of work and a household mode of production appears to
These are examples of societies in which the modes of production
are constituted (and differentiated) entirely by the organization of work and
the organization of reproduction.
are also societies in which a surplus is produced, over and above the needs or
level of consumption of immediate producers, but the transfer of surplus from
producers to consumers is organized, is dominated, by the forms of organizing
work and reproduction. Thus although surplus transfer occurs it occurs among
producers (past, present and future, e.g. surplus transfer to as yet
unproductive children or formerly productive elders). There is no surplus
extraction. Surplus is not consumed by a group, which does not participate in
collective production. Thus, although there is asocial production and
distribution of surplus there are no classes.
The Melanesian "Big Man" system described by Sahlins and others  seems to be a system of this type. In so-called "big-man" societies, (Melanesia is the locus classicus, but they are found in other parts of the world as well) ambitious individuals are able to use the prestige associated with generosity to achieve high status. A "big-man" does this by strategically giving gifts and calling in the debts he has accumulated, using the proceeds to feast his friends, kin etc, thus creating new debts which could be called in the future for another feast and so on. In the mean time, he gathers the reputation of a particularly clever and generous person. Often, in a given settlement or area there will be more than one such distributional entrepreneur, and the amount of goods distributed will be increased by the competition among them. The traditional Solomon Island "big-man" was typically male and production for presentation depended upon the ability of the entire household to produce a surplus above consumption needs. At present women are active as players in these systems as well.  Obviously a process like this would not work if it were entirely reciprocal. Why would the feast giver appear unusually generous if he merely gave to each debtor as much as the debtor gave to him? To aspire to high status one must be or become unusually productive in order to give away more than gets called in. Thus, Oliver reported that among the Siuai, on Bouganville, an ordinary family would cultivate enough taro to feed themselves and one or two pigs, but " an ambitious social climber” might raise ten or twenty pigs. More land would be brought under cultivation in order to feed the extra pigs and provide other foods for distribution at feasts.  Among the Busama people on New Guinea, a "big man"
"has to work harder than anyone else to keep up his stocks of food. The aspirant for honors cannot rest on his laurels but must go on holding feasts and piling up credits. It is acknowledged that he has to toil early and come home late - "his hands are never free from the earth and his forehead continually drips with sweat." 
ascribes the success of 'Elota, a Kwaio "big man" from the island of
Malaita in the Solomon Islands, to his willingness to work harder than most
others (a willingness shared or inculcated in his family).
"And work hard 'Elota
did - spending long hours in the fields, in the forest collecting cane or
roots for red dye, and in his house and men's house plaiting or pounding or
dyeing...So, too, his daughters and Tege [his second wife] were always weaving
baskets or armbands when they were not at work in gardens or doing domestic
chores. Production was a family affair, and when, at feasts, his wife and
daughters joined the men in presenting valuables, it was most often with
valuables they had earned themselves."
sees the "big man" system as having both political and economic
purposes of which the economic are more important. Politically, "big
men" provide a focus of organization, a nodal point of group interaction
more far reaching than kinship (in the societies where "big men" are
found. Kinship does play the same role on a larger scale in other instances.)
"Big men" provide otherwise acephalic groups with a certain
such a system motivates an intensification of production. In the absence of
external coercion most "primitive" peoples typically work only a
portion of the amount they potentially could. These are societies with a great
deal of leisure time. This is notably true for foragers.
One way to understand this involves Sahlins concept of the
domestic mode of production.
is an extremely fruitful concept even though there are serious problems with
the formulation that Sahlins presented. By generating his concepts out of
Chayanoff's studies of peasant households and generalizing immediately to
household economies in classless societies, Sahlins ignored the single most
salient characteristic of peasant households as noted by Wolf among others
namely that peasant households exist within the context of a
larger social formation which pumps surplus out. That is, as formulated,
Sahlin's concept of a domestic mode of production appears to ignore class. It
attracted some pointed criticism back in the seventies for that reason and (as
far as I know) was never fruitfully incorporated into Marxist theorizing of
classless societies, nor built upon. Nevertheless, Sahlins identified a
significant feature of some social formations, both class and classless. He
did conceptualize it powerfully if not fully. The domestic mode of production
is a characteristic feature of many social formations even if it dominates or
organizes very few. That fact needs to be addressed explored and theorized.
domestic mode of production is typical of many, if not all, classless
societies. It is organized around the household unit or domestic group
(usually, but not inevitably, the nuclear family), as a principle form of
organization of production, reproduction and consumption. At the base of the
domestic mode of production is the household economy. The household is charged
with production, with the determination of objectives and with the deployment
of labor power. The inter-relations of the household are the principal
relations of production in such a society. As Sahlins says, "the economic
is a modality of the intimate".
Decisions are taken with a view toward domestic contentment.
Production is for use, and clearly for the benefit of the direct producers.
Such a mode of production tends to produce only as much as the
producers conceive of wanting, rather than as much as they are capable of
producing. Hence, it is, in a sense, under-productive, or more accurately,
in an agricultural society the productivity of individual households may vary
depending upon such matters as the quality of the land being worked, age and
health of adult members of the household, number and age of offspring, etc.
One can document the consistent appearance of such variation in a number of
Some households seem unable to produce at a subsistence level
while other households have the capacity to produce a surplus.
`big man' system, and other forms of redistribution, provide a motive for
households with the capacity to produce above household need to do so. It
mobilizes potential labor power where it exists and puts it at the service of
those households at the receiving end of the redistribution. Thus it tends
both to equalize consumption in the face of unequal productive capacity and to
raise the general level of production to one, which "if still below the
technological capacity is above the domestic propensity."
Thus, we can see the `big man' system as a solution to a problem posed by the household organization of production. It is a way of organizing the production and distribution of surplus, which responds to, in a sense, is called into being by, the fundamental form of organizing work. The relations of surplus production are dominated by the organization of work and `solve' a problem posed by the household organization of work. This is precisely the opposite relationship from that which exists in the capitalist mode of production (or other forms of class society) where the organization of work is dominated and transformed by the demands of the organization of surplus extraction. 
CATEGORIES OF CLASSLESS AND CLASS SOCIETY
How does this help us to formulate the difference between classless and
class societies? Classless societies can be divided into two groups. First are
those in which the organization of work and of reproduction constitute the
mode of production. Work may be organized individually, domestically or
collectively. Much if not all work, most, if not all consumption tends to be
organized around the unit of reproductive activity. There is no production of
a social surplus. Second are those societies, typically agricultural, which
are, in Polanyi's terminology redistributive
Such societies typically organize work primarily on a household
basis (around the unit of reproduction). There is production of a surplus by
some households and transfer of that surplus (redistribution of the social
product among producers, past, present and future), but not surplus extraction
(coercive transfer of surplus from producers to non producers).
class society is one in which the relations of production and distribution of
surplus have been freed from the domination of the organization of work.
Indeed, they come to dominate the organization of work so far that surplus is
transferred to and consumed by a group that essentially no longer participates
in collective production except to oversee and guarantee the surplus transfer
(extraction). Thus one can say that in class societies, the relations of
surplus extraction become the dominant set of relations within the social
organization of production (mode of production).
is not a freestanding category; it is a relationship. Every mode of production
in the category "class society" has two fundamental and
characteristic classes. The differences between modes of production within
this large category consist primarily of the differences in the particular
social form of the surplus extraction relationship, e.g. slavery, tribute,
feudal dues, wage labor, etc. Secondarily, in the differences in the
organization of work and of reproduction. A slave society will operate
differently along many dimensions depending on whether it recruits its labor
force entirely through an ongoing process of enslavement or through the
encouragement of slave women bearing and raising children.
For a class society to survive over time, political, ideological
etc. factors must also exist. A mode of production does not merely consist of
the surplus transfer relationship. But all those other social relations are
subordinated to, and shaped by, their role in the production or preservation
of the class relation. The surplus extraction relation is determinant.
it is useful to step back again and expand upon a point introduced earlier,
but only in passing, the distinction between surplus transfer and surplus
extraction. This distinction, which has little relevance to the study of class
societies in isolation, is crucial to the problem of distinguishing between
class and classless forms. Thus, to delineate the difference between class and
classless forms we acquire a more tightly focused and fully explored
definition of what we mean by class. Surplus transfer was defined above as
redistribution of the social surplus among producers. Surplus extraction was
defined as coercive distribution of surplus from producers to non-producers.
First, note that it follows logically from this that surplus transfer does not
generate a class division, does not produce social classes, but rather is a
process that takes place within a group (class?). (Thus there is only one
"class" generated within such a mode of production, a one
"class" society is what we mean by classless, class being
relational.) Insofar as there is a social contradiction between producers and
consumers of surplus it is likely to be based on the producers higher status,
central role and greater authority rather than on the consumers use of goods
that they cannot produce. It is, in Mao's terms a non-antagonistic
 Surplus extraction, on the other hand, clearly involves two
distinct parties and is the fundamental social relation that distinguishes
class society. Surplus extraction is an antagonistic contradiction. Modes of
production characterized by surplus extraction always generate two typical
classes, products of the surplus extraction relationship.
This formulation proves particularly helpful in trying to theorize the transition from classless to class forms of society in history. Given that the presence of surplus from and its transfer from household to household is not in itself diagnostic of the existence of class, there appears to be a difficulty identifying the moment in history when class relations first appeared. (When the social relations of production first involved surplus extraction.) But the more complex theorization based upon both the organization of work and the extraction of surplus does appear capable of throwing some light on the problem of transition. There is not space in this paper to completely detail the process of transition in Mesopotamia, which is the context for which I have originally developed this line of thought but some of the implications need to be explored here.
the extreme centralization of both production and distribution in early Sumer
as well as the relatively complex technical division of labor and the heavy
reliance on collective work, it is impossible to assert that the emergence of
class is directly tied to the production of surplus. (In this situation, more
clearly than most, Polanyi's assertion that the surplus is a social relation
is obviously true. To find a social, non-household, surplus, one has to posit
the existence of surplus extraction, which tends to make the argument
circular.) If, however, we formulate the argument in terms of a shift from
surplus transfer to surplus extraction, with the underlying shift from the
determination by the organization of work to the determination by the
organization of surplus extraction, then we have markers that we have some
reasonable expectation of being able to find and recognize.
will be clearer if we look further at the difference between surplus transfer
and surplus extraction, concentrating on an element mentioned several times by
now but not yet analyzed. In a real sense one could say that the distinction
between transfer and extraction is political. The antagonistic contradiction
only exists in a clear and unambiguous way when the transfer of surplus is
coerced. Thus, although the concepts of surplus extraction and mode of
production were originally conceived to express a fundamentally economic
relationship they cannot be understood if we limit ourselves to the narrowly
is a commonplace among (at least) a certain school of Marxists to argue that
any mode of production must be thought of in terms of not only economic, but
also of political and ideological activity (relations). In the specifically
Althusserian version of this line of thought it is put in terms of
"levels of practice", which are "articulated", with the
economic being "determinant in the last instance". This is
accompanied by an assertion of the "relative autonomy" of the other
levels, made possible by the fact that the "last instance" never
The problem is not solved by adding a revision of the concept of
over determination, which abolishes the last instance but globalizes relative
autonomy. The real problem is that the notion of levels is fundamentally
misleading. It focuses attention in the wrong place.
implication of the theory of "levels of practice" or
"instances" (not necessarily intended), is that a mode of production
consists of, is defined by, specific and separate forms of activity both at
the level of the infrastructure, the relations of production and of the
superstructure (religion, law, art, etc.). But what needs to be recognized,
(and is by an increasing number of authors, e.g. Godelier) is that certain
of what appears to be superstructure, e.g. kinship are, in many
circumstances in fact, infrastructural. Or better, they are part of the social
relations or social organization of production. So that if we say that kinship
rules or religious practices or laws of property are part of what constitute a
particular mode of production we are not saying that the `ideological' or the
`political' are separate from the `economic' but also important.
On the contrary, we are saying that kinship rules or laws of
property constitute, expressed in what to us is the language of ideology or
politics, an essential part of the social relations of production, at the same
"level of practice" as what we regularly call the economic. I can
offer two illustrations: a negative case and a positive one. The negative case
is that of religion.
notion that religion is a key to understanding the earliest class societies is
not especially novel. It has been common to refer to the early states of
Mesopotamia as theocracies. But merely pointing to the seeming importance of
religion in the early state does not explain the transition from class to
classlessness. In a very interesting essay Godelier approached this same
conclusion, which he theorized as "the dominance of the ideological"
(using Althusser's distinction between the dominant and the determinant).
When I first began to investigate the transition I placed my
emphasis on the division of labor. Convinced of the centrality of the division
of mental and physical labor, I could explain the centrality of religion in
terms of it being the earliest example of specialized mental, or at least
intangible, work. Godelier, on the other hand was looking to religion to solve
the problem, which is a real one, of the role of "consent", as
opposed to violence, in the transition to class society. He placed great
weight on the idea that while in a technologically unsophisticated economy it
is difficult or impossible to organize a monopoly of control or access to the
means of physical production, it is easier to organize a "monopoly of the
means (for us imaginary) of reproduction of the universe and life."
Rulers could remove themselves from actual collective labor
without seeming to. Godelier formulates the first exploitation in terms of
"an Uneven exchange, more advantageous to the rulers than the ruled"
But in the Sumerian case that I am most familiar with, (and that
Godelier was also considering), the actual role of the "religious
leaders", who were really temple managers, was never as unequivocally
ideological as Godelier suggests.
further back you go into the history of Sumerian religion the less clear it is
that the ideological role was ever dominant over the managerial. According to
Thorkild Jacobsen, the earliest Sumerian deities are little more than
personifications of the means of production and the products of labor, dates,
sheep, cows grain and beer (the various versions of Dummuzzi - the product and
the `principle' of its fertility') and the storehouse itself (Inanna)
This does not imply that there was no spiritual (ideological)
component to their ritual. But the ritual clearly expresses
"economic" /managerial issues as well, even principally. Where
religion played a serious role in fourth millennium Sumerian life it did so
precisely as an expression of the relations of production. The relations
between Dama or Dumuzzi, the principal of fertility and the product and
Inanna, the temple storehouse, embodies the central organizing role in the
activities of labor and of redistribution of the 'sanga's, the temple
managers. There is not a separate and "relatively autonomous"
ideological level. The ceremonial life of the community is too intimately a
constitutive element of the organization of work and the redistribution of the
product to be considered autonomous in any way. The ritual tasks associated
with harvesting, filling the storehouse and distributing its contents were
closely interwoven with more earthly tasks or organizing production and
distribution, record keeping, and so on. Religion, in this context, simply
cannot be understood until the "economic" institutions and the role
of the temple managers ('sanga's) is itself delineated. Only then does the
religious element make sense. Only then can the "ideological"
component be explicated.
positive case that I refer to is precisely the role of the political with
which this discussion began. The initial problem here was the debate between
Emmanuel Terray and Pierre-Phillipe Rey on the question of whether or not
classes existed in what both referred to as a "lineage mode of
The particular case they were arguing over was a society in which
control over the social product seems to be in the hands, not of the producers
as a whole, but of the elders, particularly the male elders. (The question of
gender did not play an important role in their debate, nor would considering
it alter this paper's theoretical conclusions, so I will, here, leave it
aside.) Thus, at any moment, there exists a surplus appropriation, by the
elders, of the product, or some of the product, of the young men. On the other
hand, all the men have the potential of themselves becoming elders. (This is
true even though all young men do not live long enough to actually occupy that
status.) Clearly this is at least a relationship of surplus transfer. But how
to decide if it is exploitative (what kind of contradiction is it)? Terray
argued that it was classless, Rey that it had classes. Rey convinced Terray
that he was right.
Knowing too little about the particular case, I have no idea who
is correct. What fascinated me was not the question of the correct answer, but
the theoretical question of how would you decide.
In thinking about this problem, I remembered Sahlin's discussion of the breakdown of surplus transfer from productive households to the ritual chiefs on Tikopia during a severe famine.  At first, the Tikopians responded to the famine by re-emphasizing and expanding superfamilial units of consumption. Family groups that usually ate as individual households cooked and ate together. But as the famine became more severe, households pulled out of their obligations to relatives and to chiefs. The redistributive and reciprocal networks that seemed at first to be the basis of Tikopian society broke down. Households, which in normal times shared their meals with a guest of the slightest acquaintance and prided themselves on their generosity and hospitality, hid food even from close relatives. Stealing increased drastically and extended from the theft of luxury goods to theft of staples. (Thefts focused chiefly property and were rationalized by complaints about the failure of chiefs to live up to their obligations of generosity, obligations which were the mirror image, in normal times, of the traditional prestations which were now not being offered.
beyond the household held on in the formal code, but the code was being
systematically honored in the breach so that even as Tikopian society managed
a kind of moral continuity it showed itself founded on a basic
discontinuity.... "What the famine did", Firth wrote, "was to
reveal the fundamental solidarity of the elementary family".
tested structural tolerances. It exposed the weakness of that celebrated "We, the Tikopia" by the strength of the private household. The household proved a fortress of self interest which in the crisis cut itself apart. 
Sahlins interpreted this convincingly as a crise
revelatrice, which laid bare the
dominance of the domestic mode of production over the redistributive elements
of the economy. In other words, what this showed was the domination of the
organization of work and reproduction on a household basis over the
distribution of surplus. The crucial element in the Tikopia instance, which
allowed it to develop in the way in which it did, was precisely the fact that
the surplus transfers from household to household and from household to chief,
when they took place, were not coerced. This is clear because when they ceased
taking place there was no social mechanism for coercing them.
In a classless economy, even a chiefdom like
Tikopia, when it becomes necessary for the economic good of the household, the
household can withdraw from the public sphere and concentrate on its own
survival. The entire economy of a classless society, the public as well as the
family sectors, is organized for the greater good of its member households.
They may on occasion contribute more than they receive. They know that if need
arose they would receive more than they could contribute. No household can be
forced to contribute to its detriment. There are mechanisms, which encourage
and allow participation but none, other than the force of custom and social
disapproval that enforce it. Coercion in the surplus transfer relation is the
signal that the relations of surplus appropriation are more important to the
total organization of society, are dominant over, the organization of work.
Or, to put it another way, the existence of organized coercion, institutions
of coercion is/are the element(s) in the social relations of production which
"decides" that the relations of surplus appropriation have more
causal weight in determining behavior than the organization of work.
Of course, it conflates too much to talk of the
actual transition in Sumer as involving only one mode of production. It
probably makes sense to think of the household organization of work that
Sahlins describes on Tikopia as a separate mode of production from the
collective work and chiefly collection and redistribution. What we have here
are social formations in which more than one mode of production is present,
the usual, perhaps the only way things actually happen. Within such a social
formation the multiple sets of social relations constituting the several modes
of production are structured. It is, in Vilar's phrase, a "structure of
Where more than one mode of production exists, one tends to be
dominant. The dominance of one mode of production over others in a social
formation has been most usefully defined by Rey. He argues that the domination
of a mode of production means that the dominant mode of production subjects
the other modes of production to the necessities of its own reproduction, (or
minimally, that they are limited by the reproduction of the dominant mode). In
effect, this means that the dominant means of production limits, and to a
degree organizes, the scope of the dominated modes of production, either
`fencing them in', or deforming them toward its own distinct social relations
This is not a static relationship, but a process, which can be
best understood as a contradiction and thus, at least potentially, as a
conflict. This conflict, of course, occurs not between structures, but between
people behaving in socially determined ways, i.e. classes and class fractions,
which the structural relations generate.
Precisely because modes of production do not
exist as pure, separate entities, but rather embedded in real societies
(social formations), classes also do not exist in the straight forward,
unequivocal way that they can be deduced from the social organization of a
particular mode of production. The dominant class relations that we see in a
given society are primarily, but only primarily, the product of the
contradictions of the dominant mode of production. Classes exist at the level
of the social formation, which is to say, at the level of real history, not
analytical constructs. Actual social classes represent the condensation of the
effects of a number of relations, primarily the dominant contradiction of the
mode of production, which generated them but also a number of other social
relations, the product of structural contradictions both within and between
modes of production, which make up a particular social formation. Similarly,
ideology, politics, etc, as we see it expressed within a particular social
formation will be the (contradictory), result of all that is happening and not
merely the primary contradiction. Class contradiction is a structural event.
Class conflict is the historical consequence/form of structural
contradictions. Social classes, actual historical groups of people develop,
acquire/produce consciousness, culture etc in the context of the structural
contradictions of which they are the historical or phenomenal form.
On this basis we can return to the critique of
certain contemporary theorists with which this paper began. The problem with
Roemer is that he believes that by logically extracting from the historical
details of capitalism to a high enough level of logical abstraction, he
escapes from capitalism itself and generates a theory capable of dealing with
socialism. But all he really does is escape from the grubby details of
actually existing capitalist social formations with all of their deformations
from the ideal type to some historically impossible vision of a social system
in which there are no conflicts over differences of power or differences of
interest embedded in the organization of work and differential ownership of
production and the purchase and sale of labor power constitute the essence of
exploitation. What we have here, of course, is not an abstract transepochal
category of "exploitation" but an ideal type of capitalist
exploitation, precisely what Roemer believes he is getting beyond. He manages
to get the actual dynamic of historical bureaucratic socialism absolutely
backward. You might as easily argue that if you cancelled out property
relations, simply as a logical given, the way he cancels out any difficulties
in the organization of work, and if positing genuinely collective ownership in
the means of production, you organized the process of production according to
characteristically hierarchical and fordist methods, then you would reproduce
a class system based, not on ownership of the means of production, but on
control, political control over the means of production. Those who are located
by the process of history with their hands on the controls will end up in a
very different class position than those who are located by the same process
of history as the managed. As Lenin knew, even if in practice he ignored it,
the division of mental and physical labor, the dialectic of manager and
managed is in and of itself capable of (re) generating a class contradiction.
In a real sense, that is what happened in the
course of the pristine emergence of a class contradiction, in ancient
Mesopotamia, where, in all likelihood, the first class society emerged in
Sumer sometime around the fourth millennium B.C. The course of development of
that region over the previous six thousand years can be summed up briefly. The
emergence of class society in Sumer is the heir to a process, which began with
the development of agriculture out of foraging. Agriculture emerged in
Mesopotamia as a process of domestication of wild species of grains and other
plants and animals on the hilly flanks of Mesopotamia where the ancestral wild
species of wheat, barley, sheep and goats appear to have been domesticated for
the first time. This is not the place to speculate why foraging people made
the shift to horticulture, but the archeology makes it clear that the earliest
farmers were, for all intents and purposes, as egalitarian as the hunting and
gathering bands they descended from. As people moved from the hilly flanks of
the region into the river valleys proper, certain innovations were necessary
for the agricultural villages to sustain themselves on the basis of the
existing patterns of planting and eating. Three changes, at least, were
required. Two of these might be perceived as technical; one is clearly a
matter of social organization. They are, both logically and in the record,
profoundly interdependent. The first change is the development of irrigation
agriculture. We are talking about people moving from regions of frequent
rainfall to regions where water, while available, was not readily available
exactly where it was needed. The second change is the invention of
transhumance. As they moved out of the hills into, where it is not irrigated,
a more arid landscape, they needed to find new food for their flocks of sheep
and goats. Rather than devote themselves to the year round production of
fodder and unwilling to let the beasts run wild in their fields, it made sense
to take them back up into the hills where there was ample pasturage during the
growing season, then bring them back when the harvest was over to let them
graze on the stubble in the fields. So the pattern of stock raising called
transhumance, moving the flocks back and forth between winter and summer
pastures, became established in and around the same time as irrigation
agriculture, in response, undoubtedly, to some of the same material
conditions. Irrigation made it possible for people to live in certain places.
With irrigation the crop yields were by ancient standards impressive. Living
in those places made transhumance if not inevitable, at least attractive.
The problem, of course, with both irrigation
agriculture and transhumance if they both going on this way is that they
interrupt what had been up until then the normal pattern of farming whereby
each household appears to have taken care of both planting and herding,
reproducing within a horticultural lifestyle, the household integration of the
economy that characterized hunting and gathering bands. There may well have
been some division of labor within the households - but the social product got
redistributed around the hearth in the process of cooking and consuming the
meal. But irrigation agriculture and transhumance created a situation in which
the herdsmen were separating themselves by considerable distance from the
farmers. We are talking about the emergence of a socially significant division
of labor, a division of labor which threatened to deprive the herders for a
significant period of time of the produce of the farmers and the farmers for a
considerable part of the year of the products of the herds.
How were those who were no longer raising their own animals to
provision themselves with milk, meat and cheese? How were those who were
herding at some considerable distance from the fields to provision themselves
with pilaf or porridge or bread or beer? The emergence of the division of
labor as a social fact within the community clearly occasioned the need, if it
was going to survive as a socially and economically integrated community, for
a novel mechanism for distribution of the social product. These communities
appear to have developed one in the form of a communal storehouse and a ration
system whereby the social product beyond the immediate needs of the direct
producers was stored in a community institution from whose stores members of
each occupational group could draw for their supplies.
The social mechanisms which served to integrate
stationary farmers and transhumant herders were easily adaptable to
specialization of other kinds, e.g. full time fishing. In early historical
times, most towns in Sumer (the southernmost part of Mesopotamia) had at least
two separate groups of full time fishermen, freshwater and salt water from the
sometimes three (rivers and lakes).
The adaptation of the existing system of specialized labor and
redistribution must have encouraged the development of full time fishing
almost as soon as the early neolithic villagers settled in the river valleys.
By early historical times the division of labor
was remarkably complex. In addition to basic subsistence work, farming (of
wheat, barley, legumes, sesame for oil, dates, other fruits and vegetables),
herding (sheep and goats for meat and wool, cattle for meat milk and hides),
fishing (fresh and salt water), and birding there were numerous other crafts
each organized as a clan (Sumerian im-ru-a). There were im-ru-a specializing
in pottery, brick-making, carpentry, beer making, dairying, weaving,
upholstering, copper and silver work and many other highly specialized craft
and ritual activities.
Most of the agriculture took place on land "owned" by
the agricultural im-ru-a, but there was also land which "belonged"
directly to the sanctuary-storehouses. The "temple land" and the kin
group land did not belong to separate spheres. In some cases 'temple land was
share cropped. The archives of Innin at Ur show "land, draft animals and
seed were turned over to community members for cultivation and the temple got
a share of the crop"
Some work in the "public" sector was
done by workers who farmed their own land and also worked on temple land or on
other public projects. Farmers were organized to do public works,
construction, canal building and repair etc for portions of the year.
Fishermen and workers who preserved fish worked directly for the
temple storehouses. 
The herds were considered to belong to the temple
and were milked in centralized milking parlors
Since the temples owned the herds, it followed that they owned the
draft animals. As a result, plowing was centrally organized by the temples
which provided teams of plowmen to the im-ru-a at planting time.
Documents from Ur speak of a "house of the plows" which
distributed plows and draft animals
The temples were storehouses which collected and redistributed
both food and seed grain.
The Sumerian word "ensi" which in latter Mesopotamian
history meant the ruler or governor of a town originally referred to a temple
connected leader who organized and supervised agricultural work.
The chief manager of a temple storehouse was referred to as
 The temples were central in the supervision, organization and
coordination of production, in the collection and redistribution of labor and
the products of labor. The temple storehouses were also the centers of long
distance trade Leemans 1950
for items like stone (flint and obsidian) and latter copper, which
were required for tool making and for lumber, which was rare in the river
valleys. They supervised the shearing of sheep.
Weaving was organized in temple connected workshops which provided
the clothing worn by ordinary workers and the cloth which was traded for
stone, copper etc.
Work was organized and carried out collectively
by work groups, supervised by "foremen" who in the early period were
kin group leaders.
Some of the work teams were full time, e.g. weavers, smiths,
fishermen and those who cured fish and long distance traders. Others were part
time workers like the plowmen or the corvee workers called upon to do seasonal
public works. When organized military groups appear in the early Dynastic
Period these im-ru-a, the kin based craft and work organizations under family
leaders, became the units of the army. Military organization was translated
directly from the organization of peacetime work.
Workers on public projects received rations, principally barley
and oil but also beer, bread, fish, dates, peas, etc. The also received
periodic allocations of cloth and garments.
Farmers kept a portion of what they grew for their own subsistence
and received what they did not produce in rations from the storehouses.
Herders received rations in their mountain camps, similar to those
of the corvee workers.
The increase in the importance of collective
work increased the importance of the management and coordination of that work.
The integration of production and distribution through the ration system
required even more managerial activity. The combined management of production
and distribution became a major form of socially necessary labor. The 'ensi's
and 'sanga's running the temple storehouse ration system and the work teams
invented writing to help them keep track of the inputs and outputs. Managing
productive work became a separate activity from doing productive work. Mental
labor was separated from manual labor for the first time. This is not yet
exploitation in a pure sense. Managing is socially necessary labor. Collective
labor on this scale required coordination. The work of the 'ensi's and the
'sanga's was a part of collective production.
Nevertheless, something quite new had happened.
A group within society had emerged who lived off of the surplus created by the
physical labor of others. And that group was in charge. A great many of the
ideological, social psychological or cultural features that one associates
with class society were not yet present. The hundreds or thousands of years of
history in class society have not yet produced them. They would come. The
fundamental structural relation had developed for the first time.
Thus was born a contradiction, which was not,
itself, a clear contradiction between classes but certainly constituted the
next closest thing. The development of the division of labor, and the
embodiment in the relations of production of supervision and subordination
directly related to the division of labor, did not by itself constitute
antagonistic classes, but it brought society to the brink. As these relations
were supplemented and consolidated by the forms of judicial and political
coercion familiar from the ancient law codes,
the social relations of subordination and surplus transfer to the
redistributive managers hardened into true surplus appropriation, that is to
say, class society.
Thus we see that in at least one case, a case
which is directly ancestral to our own society and whose particular
development and dynamics are still significant for the course of western
it was the development of the division of labor and not the logic
of property relations which brought antagonistic social classes into being.
Indeed, at that early stage of Sumerian society one could hardly say that
private property in the means of production existed at all. Concepts of
ownership, such as they were, developed in Sumer in the context of the
subsequent struggle between the old leaders of the im-ru-a, the kin group
elders, and the new rulers of society, the 'sanga's and 'en's of the temple
storehouses. With the emergence of class contradiction, that is, we also see
the emergence of class conflict. As in every period, conflict between classes
and sections of classes altered the historical configuration, producing the
particular groupings and awareness of social structure that we see.
So, it is not surprising that the classes, which
we find in the Mesopotamian record for the Early Dynastic period and later,
are somewhat differently constituted than the class configuration of the
period of transition from classless to class society itself. The classes that
we know from contemporaneous written records are already the product of a long
period of struggle. Not merely struggle between the dominant group of economic
managers who are usually identified as scribes or priests and the variety of
laborers, but also conflict between the scribes/temple managers/priests and
the hereditary leaders of the corporate kin-groups, im-ru-a). The kin-group
leaders appear to have struggled, with mixed success, to avoid being reduced
to the level of intermediate managers, foremen or subalterns. It is quite
clear that in some cases kin-group leaders were reduced to some such status.
It is also clear from documents such as "The Reforms of Urukagina"
that as 'foremen' they had a tendency to assert greater degrees of control,
even "ownership', over the assets, whether it was that which was produced
or that which was redistributed, than popular tradition assigned to them.
Class struggle between laborers and foremen, the "chief of the
boatmen" or the "chief of the herdsmen" was one of the issues
in society the factions within the dominant group could use to advance their
own interests against other factions of the dominant group. (This is a major
theme of the "reforms of Urukagina".)
Other leaders of the im-ru-a seem to have
successfully transformed their positions as kin-group leaders and therefore
managers of labor into an early, although partial, form of private property in
the kin-groups land, which allowed them, apparently, to assert a degree of
equality with the temple managers. This is the origin of the
"estates" which play a major role in organizing agricultural
production in Post Sargonic Mesopotamia
Some, at least, of the leaders of the agricultural im-ru-a managed
to transform clan land into personal property and become themselves
small-scale managers. They then used their traditional authority as members of
the council of elders to contest the preeminence of the temple-managers. As
populations grew and conflict between towns over agricultural land along the
borders became common, the kinship organization of the military gave elders a
political advantage. It was a prerogative of the elders to elect a
"lugal" (literally big man)
to serve temporarily as war leader in times of need. (Some similar
arrangement institution appears to have been preserved in the Roman office of
dictator.) The elders appear to have used their authority to elect a lugal as
a way of asserting collective (i.e. class) political power against the scribes
and temple managers. Evidence suggests that the lugals were well rewarded for
this function and also that they were not, at least in known cases, themselves
owners of property.
This is suggested by two facts. Many, if not most, bills of sale
of land surviving from this period relate to the purchase of property by
'lugal's from groups of clansmen.
Clearly, the lugals were engaged in setting themselves up as
landed proprietors on the model of their patrons. We also can see, in the
reforms of Urukagina and other documents, evidence that the lugals and their
families tried, with varying degrees of success, to treat the collective
property of the "temple storehouse" or the city as if it was their
As a result, by historical times, the relative
power of the temple managers, the first ruling class of the period of the
transition, was already circumscribed and a new class, emerging in reaction to
the existence of the temple managers, was on the rise. Further, it had evolved
from an original form in which land ownership was totally encumbered by
recognizable subsidiary rights of fellow clansmen, into a second phase where
those rights had, at least in some cases, been extinguished by purchase and
perhaps other means. Only at this point could the creditor -debtor
relationship, which Diakonoff, following Engels, takes to be the critical
motor of class formation, have come seriously into play.
In the struggle between the temple managers and
the elders, the emergence of "monarchy "(the permanent lugal) as an
element of this struggle creating new forms of political control over the
economic-ideological institutions of the temple storehouse and its workshops,
over the ration system etc, there was a long period when institutional control
slid first one way then the other in each town. The struggle between these two
groups is a major dynamic of the entire Early Dynastic period.
Economic institutions, which had originally
emerged under the political control of the council of elders,
had gradually escaped their control as the importance and
therefore the power of the temple administrators increased. This was the
period in which ideological, economic and political authority were vested in
an official called "en", who, in this period, headed the hierarchy
of temple administrators.
His power was a product of the emerging contradiction between
mental and physical labor which produced the first class contradiction.
Apparently, the emergence of full time management brought with it an increase
in the quantity and quality of labor discipline, i.e. coercion. This
developing practice of coercion constitutes the emergence of the state in the
Weberian sense of the monopolizer of legitimate violence and thus established
the scribes and temple managers as a dominant class.
But the very need for coercion which empowered
them, made their uncontested sole dominance unlikely. The control of justice
and of military power remained in the hands of the council of elders.
This gave them the ability to reassert their authority within the
state. This led eventually to the emergence of hereditary regimes of lugals or
ensis (the term used in this period in some cities as an alternative term for
a non-priestly head of state) and a contest for control over both the state
and the economy between these 'political figures' and the temple managers.
Structurally, what happened was, the
incorporation of the older im-ru-a based egalitarian mode of production into
the social formation dominated by the new temple storehouse mode of
production. Then, the subordination and alteration of the im-ru-a mode of
production by the temple storehouse mode of production and finally the
modification of the dominant temple storehouse mode of production through
That this conflict occurred, how it was enabled by the struggle for power and benefits among the subordinate class, is exemplified in the text known as "The Reforms of Urukagina" and similar documents produced during the reign of Urukagina as Ensi of Lagash, circa 2400 B.C.  The documents are full of complaints of ordinary workers against the priest-scribes, and against the foremen under whom they worked, complaints of the priests against the former ensi, lugalubanda, for appropriating control over the temple economy and against the foremen for "embezzling" product which ought to have been turned over to the workers or deposited in the temple storehouse. These grievances constituted the fuel for a "coup d'etat" which put Urukagina in the ensi's office replacing the much disliked Lugalubanda.  We lack similar detailed descriptions from every city in Sumer, but this conflict between elders and priest managers, workers and foremen, priest managers and political rulers, fueled unrest in each city and must have contributed to the instability of governments and brief tenure of rulers, conflicts whose details may be unknown but whose outline is clear in the documents which have survived.
what we have seen by examining the Sumerian case is that at the moment of
birth of class society (In one of the half dozen or so pristine cases) the
emergence of social class and its historical elaboration are intimately linked
with the emergence and elaboration of the state. Just as it was control over
the state, which constituted the first class contradiction, it was control
over the state, which constituted the prize in the struggle between im-ru-a
elders and the temple managers. The emergence of the state is the phenomenal
form of the first class contradiction. Conflict over control of the state is
the phenomenal form of the first well-documented class conflict. Because the
emergence of class was contingent on the emergence of the state, because the
state was constitutive of the emergent class relations, the newly emergent
state was the nodal point at which the stability of the social formation could
be contested or protected. Only from a narrowly economistic point of view does
it make sense to analyse this process in terms of the Althusserian three
levels. What we think of, as elements of the political and ideological are too
thoroughly determinant of the relations of production. The emergence of
political control over the organization of work is precisely the historical
change, which makes surplus appropriation historically possible. I might add
here that while I am suspicious of rigidly generalizing claims about stages of
history, Jonathan Haas work on the emergence of the state in South America
suggests that there too, the state was the phenomenal form of the development
of class conflict.
We have been mislead in our theorizing, to some degree because of
our understandable focus upon capitalist social formations. Power, in
capitalist societies is clearly diffused between political and the economic
institutions. The theory of levels appears to make sense as analysis of
capitalist social relations. But we need a different angle of vision to
theorize the possibility of the transition from capitalism to socialism. That
angle of vision must recognize that elements of what we have called the
political are as central to the social organization of production in any class
society as the elements that we have called economic. Marxists have wrestled
with versions of this understanding all along, in the attempts to theorize the
role of the state in a socialist transformation, in counterpoising state
ownership to private, or in recognizing the political nature of revolution,
but it has not been adequately conceptualized and the consequences, not merely
for theory, have been vile.
A final consequence of this line of thought concerns the relationship between the politics of class and the politics of democracy. If, as I have been arguing, the social organization of production rests on the meeting point of exploitation and coercion then struggles over the distribution of the social product, struggles about control over the organization of work, and struggles over control of the organization of the state are not merely parallel goods, equally ethical, equally desirable. They are struggles, from different historical angles of vision against the same fundamental transepochal circumstance that created what we know as class society. No victory in one or two of these areas would be sufficient to abolish class society, but only to transform it into yet another of its many historical forms.
[*] An earlier version of this paper was read at the Marxism Now: traditions and difference Conference, UMass Amherst, Amherst MA, Dec.1, 1989. I would especially like to thank Peggy Somers, Gerry Gamburd and Elise Brenner for their comments on earlier versions of this paper.
 Roemer 1982, Roemer 1988, Roemer 1989
 Roemer, 1989, p 3
 ibid, p99
ibid p 93
 Buroway 1979,; Buroway 1985; Buroway 1989a, "Should we give up on socialism" SR v.19,no.1 pp59-76 January,1989b "Reflections on the class consciousness of Hungarian steelworkers" Poitics and Society v 17 n 1 pp1-34; 1989c "Marxism without microfoundations" SR v19 n2 pp53-111
 Buroway 1989c p 80
 Laclau and Mouffe, 1985, Hegemony and Socialist Strategy, Verso, Bowles and Gintis, 1986, Democracy and Capitalism
 Resnick and Wolfe, 1987, Knowledge and class, p
 Therborn 1976
 Cohen, G.A. Karl Marx's Theory of History Princeton 1978
 Turnbull 1961, Turnbull 1963, Turnbull 1983
 Lee 1979, Lee 1984, Marshall and Ritchie 1984
 Sahlins, 1963, Pospisil, 1978 The Kapauku Papuans of West New Guinea
Keesing 1983 p124
Oliver, 1949, p89
Hogbin 1951 p131
Keesing, 1983, p9
 Lee 1979,pp250-280, Godelier 1975, Sahlins 1972a
 Sahlins 1972a, pp??.
Wolf 19?? Peasants
Sahlins 1972a, p
 c.f. Bravernan 1974, Marglin 1974, Stone 1974
 see Weber, The Agrarian Sociology of Ancient Civilizations and Perry Anderson From Antiquity to Feudalism ????
 c.f. Althusser, 1969, p94
 Poulantzas 1972 pp70-71, Fioravanti 1972 pp53-54, Terray 1975
 Godelier would say function, but that seems too loaded a usage. Also Godelier 1986 pp208-226
 Resnick and Wolfe, 1987,p
Godelier 1978, Althusser 1970
Godelier 1978 p94
Jacobsen 1976 pp23-75
 Terray ,1972, Terray 1975, Rey 1975
 Sahlins 1972b
 Firth 1959 quoted by Sahlins 1972 p128
 Sahlins op.cit. p129
 Pierre Vilar 1973 p?
 Pierre- Phillipe Rey 1973, p ? Foster Carter 1978
 Butzer, 1964, Haelback, 1964, Flannery 1965, Flannery 1969, Flannery Hole, and Neely 1969, Hole, 1977, for a more detailed discussion of this material see Miller 1980, chapter 3.
 Flannery 1965, Johnson 1973, Miller 1980
 Miller 1980, Wright 1969,Gelb 1965, Diemel, 1931
 Van Buren 1948, p102, Gadd 1964 p 36
 Lloyd and Safar 1947 p103
1966, Wright 1969
Tyumenev 1956 p72
ibid p 71
 Gadd 1962
 Adams 1955
 Mallowan 1965 p17
 Tyumenev op.cit. p74
 Saggs 1962 p44
 Tyumenev op.cit. p77
 Tyumenev op cit p 72
 Ibid p75
 Adams op cit
 Wright 1969 p 41
 Alfred Crosby, Ecological Imperialism, the biological expansion of Europe 900-1900,Cambridge, 1986, C.U.P.
 Lambert 1956, Diakonoff 1958
 Diakonoff 1967, Gelb, 1965
 Jacobson 1957
 Adams 1966 p 140
 Diakonoff 1954
 Diakonoff 1958
 Diakonoff 1968
 Tyumenev op.cit
 Adams , op cit p 128, p137, Jacobsen and Kramer, 1953 p181
 Jacobsen op cit
 Lambert 1956, Diakonoff 1958, Gadd, 1962
 Gadd op cit
 Haas 1982, The Evolution of the Prehistoric State, Columbia U. Press
Partly edited for this Website by Chambis Kiatipis