Marxism is not simply a form of anti-capitalism, but is a theory of the social and historical nature of humanity. Capitalism was understood by Marx to be a specific historical phase, in which the cycles and tendencies of human activity are qualitatively different than in previous periods, each of which had their own specificity. Capitalism does not represent merely a quantitative increase in a general category of human activity, such as ‘industry’, or ‘the market’, when the latter is conceived as a historically undifferentiated mechanism of exchange. Nor does Marxism accept forms of determinist thinking that rely on single factors, reducing society to its dependence on say, stone, iron or oil, still less on unmediated tendencies, such as population growth, as found in Malthusian theories, for example.
Capitalism is, therefore, a ‘mode of production’, a complex system composed of the social forces of production (in capitalism this would include the organisation and technology of industrial production) around which are articulated specific social relations of production (in capitalism, this would include ‘the market’). The operation of the forces and relations of production require certain ‘superstructural’ institutions, political ones for example, to ensure their reproduction in their existing forms. This is because human social systems have been, so far in history, necessarily contradictory; ‘the history of all hitherto existing society is the history of class struggles’.
Our present focus, as socialists, tends to be on the present, and the class struggle under capitalism. Nonetheless, the nature of pre-capitalist systems must be of serious significance to Marxists. If capitalism really is just a historical phase, and if Marxist explanations of politics in terms of socio-economic structures (forces and relations of production) are to hold true, then Marxism as a historical method, as well as a politics, needs to be able to explain, convincingly, how social developments arrived at capitalism.
Doing that requires an analysis of pre-capitalist modes of production, the foundations of which were laid by Marx and Engels. This area has not been the least controversial subject of debate about Marxist theory by any means. This new collection of articles in the Historical Materialism series is therefore very welcome as it brings to bear a varied range of case studies from the ancient world to the medieval period on many of the theoretical issues that have haunted the debate over the years.
Stages and modes of production
The disagreements are indeed substantial and can seem intractable at a basic level, given that there is not even any consensus about the applicability of the different types of pre-capitalist modes of production identified by Marx. Stalinist historical theory reduced a plurality of different possible modes to a single rigid sequence for all class societies, going from the ancient slave mode, through feudalism to capitalism. This formula has since been largely rejected for a number of reasons, including its mechanical determinism, but it was also problematic as it tended to wipe out the clearly observable differences between the many distinct pre-capitalist class societies on all the continents.
The rejection of Stalinist mechanical determinism has done little to resolve the underlying debates. Some historians, such as Chris Wickham, who features in this collection, argue that the only distinction that can be made between pre-capitalist class societies is that between one based on a slave mode of production and all the others, based on lord-tenant relations, or state taxation regimes. These latter are sometimes all summed up as the ‘tributary mode of production’ (subsuming European feudalism, most ancient states, and all other class societies from India to China, for example).
John Haldon and others go further, and see it as only possible to conceptualise one mode for pre-capitalist class societies. This would reduce the total number of possible modes of production before socialism to three; pre-class society, pre-capitalist class society, and capitalism. Others do not accept this reductionism. In this collection, Andrea Zingarelli makes a strong case for the applicability of Marx’s ‘Asiatic mode of production’ to the case of the ancient Egyptian state, in all three of its Old Kingdom, Middle Kingdom and New Kingdom iterations. The primacy of the state inhibits the development of private property in this system, and sets up distinctive social relations as a result. However, historians like Haldon would presumably respond by denying that an analytical distinction between state and private property can differentiate past modes of production.
The law of value in history
A resolution to all these problems will not be made soon, but this collection contains some important arguments, and helps to clarify the parameters of the debate. Perhaps one of the most illuminating contributions adopts an apparently limited focus by investigating the applicability of the law of value to pre-capitalist economies. Octavio Colombo begins with a few apparently contradictory statements by Marx and Engels on whether value as determined by labour time governs all systems of exchange, capitalist and pre-capitalist, or not. From one view, given certain conditions, it did apply to the produce of craftsmen and peasants alike ‘in the ancient as well as the modern world’ (p.239). In a different analysis, it is only the development of the capitalist market which enables the law of value to operate, whereas beforehand it did so only ‘in a qualitative sense’, or only in incomplete and imperfects ways (p.240).
The law of value, Colombo explains, requires certain conditions for ‘the tendential adaptation of prices to values’ to take place consistently (p.266). That is to say, the value of a particular commodity, which is the amount of social labour time expended upon it, requires the right social conditions for the price, the amount for which it is actually exchanged, to tend to reflect that value, or socially necessary labour time. For example, capital needs to be able to move freely between sectors of production, in order to find the most profitable investment. In addition, there needs to be a sufficient consistency and completeness of exchange relations, for commodities to be regularly exchanged according to abstract exchange values. In fact, before fully capitalist relations were established, peasant produce could often be exchanged at less than its labour value, if peasants were simply selling their surplus beyond their subsistence needs; ‘Under these conditions, peasant surplus labour time is given to society for free’ (p.240).
The way this works is that the peasant family produces its own subsistence, but will do the extra work to produce a surplus to sell in local markets, possibly to earn the money to pay taxes or rent demanded in cash, or even for a small discretionary income above subsistence. In these circumstances, peasants will often be willing, or more often lack the choice to refuse, to exchange their surplus for substantially less than the value of its labour. Capitalists, in contrast cannot do this consistently and still make a profit. In this way, the forms of exploitation of the peasantry before capitalism acted in such a way as to frustrate the tendential emergence of the law of value.
There are many other conditions which prevent the law of value from operating under pre-capitalist simple commodity production, not least the vagaries of nature. Where there is not sufficient social and technological control over nature’s unpredictability, the price of agricultural produce is affected frequently by dearth and over abundance. Colombo notes that feudal lords were able to take advantage of these circumstances in various ways to extract surplus from the peasantry, incidentally once more frustrating the operation of the law of value. Another problem of general, although varying, importance in pre-capitalist societies is ‘the lack of efficient means of transportation that would have allowed for the relative equalisation of the local or regional imbalances’ (p.251). Thus, while labour was ultimately the source of value still, the law of value did not operate with any consistency before the capitalist market became sufficiently dominant to condition all production and exchange in a given society.
Capitalism represents more than just a quantitative increase in production for the market by this account. Moreover, it can only emerge at a threshold moment when the law of value comes into operation over a whole social system. Colombo concludes:
‘value as a social relation can only function when prices reflect their production conditions, and this occurs only when the relations of production are governed by the law of value. In consequence, a general formulation of the concept of simple commodity production cannot be immediately applied to the study of the medieval peasant economy; first it must determine through historical analysis the manner in which the latter is shaped by the dominant mode of production’ (p.267).
This argument has important consequences for the question of the analysis of modes of production in general, because it demands an understanding of the operation of elements of society in terms of the whole. The status of simple commodity production, in this instance, cannot be determined by its own characteristics as such, but rather only through an understanding of how it is embedded within a totality. The whole determines the parts, and not the other way around.
Peasants and lords in medieval Spain
A similar perception is at work in Laura da Graca’s consideration of the feudalisation of northern Spain in the eleventh and twelfth centuries. She makes critical use of Chris Wickham’s concept of a ranked peasant society, or peasant mode of production, which he explored in Framing the Early Middle Ages (2005). This idea was certainly fruitful there in making sense of a complex and ranked society like early medieval Ireland, where a true aristocratic class was very slow to develop. Da Graca’s analysis similarly takes as its starting point an essentially egalitarian peasant society, with private property, kings and nobility, but without fully developed class relations.
Da Graca convincingly shows how peasant-based client and patron connections, with reciprocal gift relations, have their own logic, opposed to that operating through aristocratic attempts to create seigneurial relations. The peasants’ behaviour had an ability to disrupt and impede efforts at feudalisation, not least through Spanish custom whereby freemen could make their own choice of lord each year (p.195). Nonetheless, in the end it was vulnerable to aristocratic invasion such that at a certain point there was ‘a catastrophe-flip from a peasant to a feudal economic logic’ (p.169, quoting Wickham, Framing the Early Middle Ages).
However, da Graca, while showing the explanatory value of Wickham’s peasant mode, is also critical of his wider conceptual approach, arguing that it is not adequate to typify societies strictly according to ‘exploitative modes of production’ which ‘are reduced to forms of surplus extraction’. Instead, she argues that in her analysis of the feudalisation of Spanish society:
‘It follows that productive forces, and even relations of production, do not singularise a mode of production. A mode of production, according to the Formen [Marx’s writings on the subject], implies a form of property at its base … These elements [the laws of the mode of production] were assessed on the analytic level of the social formation, since they allow for the characterisation of the forms of articulation between the aristocracy and free communities, and in our case highlight the process of subordination of the latter to the feudal mode of production’ (pp.202-3).
It is possible to go further here, and suggest that the gift and client relations that da Graca discusses can, in fact, be thought of as part of the relations of production in a wider sense. These customs are in practice relations of distribution, which, as part of the social totality, form part of the social conditions for production relations.
Similarly, there is also a problem with Wickham’s characterisation of early medieval Europe, in da Graca’s words, as a picture of feudal enclaves like ‘leopard spots’ within a larger number of peasant-mode communities. The issue is not precisely this image, which is probably a good empirical description of much of the post-Roman West, but rather the conceptual consequences. It suggests that ‘modes of production’ can exist as small enclaves side by side. Jairus Banaji has pointed out that this confuses Marx’s two usages of the term. The simple meaning refers to the different specific mannersof production (factory versus handicraft, for example), which can easily exist side by side. The more complex meaning refers to the whole social system and its productive relations. In confusing these two usages, an analysis based on dialectical totality is replaced with one based on positivist categories.
Feudalism’s historical context
Assessing European feudalism in terms of ahistorical categories also does not fully account for the historical situation; the centuries after the collapse of the Roman Empire. Here the mode of production was clearly conditioned at the level of the totality by a tax collecting state, whose legions not only ensured those revenues were raised, but also guaranteed the far-flung estates of the slave-owning senatorial aristocracy. The collapse of this state, its armies, and the commodity exchange system which it supported, meant that social relations became radically localised. The Roman class structure, which had depended upon a widespread system of enforcement, could no longer be sustained. All sorts of elements of class society fell apart with the fall of the state, and the surviving aristocracy necessarily transformed itself into more regional feudal elites, with a much narrower and poorer base of exploitative relations. This is the context for La Graca’s more theoretical observation that the aristocracy ‘assimilates itself to the peasant mode’, that is to say, the old elites become merely tribal leaders (p.166). In places, for example most of Roman Britain, class society did indeed collapse altogether.
Even so, what occurs after the collapse of the Roman system is not simply an involution back to pre-Roman society, probably even in Britain. The extent to which communal and tribal property systems had been broken up would have been a key legacy. Where there is evidence, a preponderance of private peasant property appears to be normal, which Wickham took as the starting point for his peasant mode. In da Graca’s Spain, there was also the post-Roman legacy of the Visigothic successor state, as well as the later Arabic civilisation. These class systems created the conditions for the survival of an aristocracy in the northern parts of Spain.
Thus, while the ‘leopard spots’ image is usefully descriptive, it doesn’t quite capture the dynamic of a feudal mode of production in its early stage, where exploitative relations had only a limited purchase outside the great aristocratic estates that survived the Roman collapse. Rather than there being separate ‘peasant’ and ‘feudal’ modes co-existing, it would make more sense to focus on how the aristocracy used the remains of Roman political institutions (and particularly the Catholic Church) to reconstruct quasi-state forms that would guarantee their dominance over the whole society. The contrast between aristocratic and peasant modes therefore captures a picture of the nature of class struggle in the early medieval period, rather than two separate, competing modes of production. Again, the totality needs to be considered if the dynamics of the parts are to make sense.
Chris Wickham’s own contribution to this collection, an analysis of saga evidence that reveals the emerging power of a feudal aristocracy in eleventh and twelfth-century Norway, in effect also documents a class struggle parallel to that found in northern Spain. The client-patron relationships are quite different in terms of cultural detail, but bear strong structural affinities nonetheless. This essay, by the way, is a tour de force in the creative use of dubious narrative evidence. The sagas can in no way be considered reliable evidence of the events they describe, but the social assumptions of their authors are, as Wickham shows, a marvellous store of evidence for the development of social relations.
Slavery as a mode production
Nonetheless, the analysis brings up another recurring issue in these debates, which is the role of slavery in different modes of production. Slavery, like in fact wage labour, is a form of exploitation that re-appears in many different societies. While in capitalism, wage labour is the dominant form of exploitation, elsewhere the presence of wage labour as a simple form, is not indicative of the wider set of capitalist relations. Wage labour on its own account is not a sign of capitalism, in embryo or in any other sense. Similarly, the presence of slaves in a social formation is not automatically diagnostic of a slave mode of production.
Thus, in medieval Norway, the aristocracy was engaged in using client networks to gain purchase on political power, and to feudalise society, very gradually. Yet, according to Wickham, the aristocracy’s power was based on the presence of slaves on their estates. Unusually, these slaves appeared to live in the same household as their lord (pp.155-6). This would seem to indicate that far from these arrangements representing an embryonic form of the slave mode of production, slavery here was conforming to the wider household-based relations of production. Wickham indeed notes that ‘this mode [slavery], even more than tenancy, was dominated by the hegemonic structures and economic logic of the peasant mode’ (p.155).
Conversely, it is important to recognise when slavery really does condition the entire mode of production, as in the Roman Empire. Carlos García Mac Gaw takes, in contrast, a particularly reductionist approach to social relations in his analysis of slavery, insisting that as a form of exploitation, it be identified very narrowly only with work gangs on plantations (p.96). In this way, the analysis can find that slavery was only a subsidiary part of the Roman economy, particularly as slavery outside the strictly plantation context is dismissed as merely legal in nature (p.104). This enables Rome to be assimilated to the tributary mode of production (p.108).
This direction of analysis seems very unhelpful. It means that Marxism is not able to make any serious distinctions between different class formations before capitalism, if the Roman Empire cannot be qualitatively distinguished in form from the feudal society that followed it. The unfortunate results of this kind of approach seem to be confirmed by the contribution from John Haldon, where he claims point blank that ‘in the period between the second or third century and the thirteenth or fourteenth century’, there was no change of mode of production at all. This is despite the list he gives of developments in productive and military technology, trade and commerce, and state organisation, across the period. And yet:
‘once again it is impossible to see the quantitative and qualitative transformation in the forces of production that promoted the sort of modal shift required by the classic understanding of the concept “mode of production”. There were changes, shifts … none of them qualifies as a modal transformation’ (p.215).
A non-Marxist, reading this judgement, would surely conclude that the concept of a ‘mode of production’ had no analytical use or value, if it cannot grasp as significant the massive transformations across this thousand-year period in western Europe.
In contrast, if the analysis of forms of exploitation is located within an understanding of the social totality, it is easy to see that slavery in the Roman Empire was the dominant relation which tended to condition the others. It is the opposite case compared to Wickham’s picture of early medieval Norway. Slavery cannot be dismissed as a legal form in the ancient world, since those legal relations related very meaningfully to the whole set of social relations.
Failing to see the centrality of slavery in the ancient world also makes a nonsense of the patterns of historical development in that time: slavery was not an exceptional or marginal phenomenon which only briefly dominated society. Roman slavery was the culmination of a dynamic system which clearly originated in ancient Greece, notably in Athens, and was then exported to the Greek colony cities, for example, Syracuse in Sicily (Sicily, of course, was to become a key heartland of the Roman plantation system). The Roman Empire represents the maturing of a web of social relations which had slavery at its core. The Roman state was the outgrowth of this system, which existed in order to maintain slavery as the dominant relation. Of course, the state also, contradictorily, engendered processes which ultimately undermined the system of slavery and the state itself.
The medieval social order which followed had entirely different dynamics, and was unable to support anything like the Roman state. This fact ought to point a Marxist analysis towards discovering fundamentally different social relations. Moreover, the apparent regression in sophistication in the medieval period actually masks the development in efficiency of class exploitation. The Roman Empire needed a centralised system to concentrate the fruits of exploitation from over a very wide area, and the result was the appearance of fabulous power and wealth in a few centres.
Conversely, medieval states were slow to develop any great power or geographic range, partly because networks of landowners were able to maintain class relations on a much more local level, without massive state assistance. This can only be because individual property holding was much more embedded than it had been before the Roman Empire. Feudalism as a mode of production reflects the development of social forces of production by the preceding mode. Assuming both social formations are identical in depending upon tenancy relations (with taxation folded into that category) flattens out observable historical development, and effectively abandons the attempt to explain it through the Marxist method.
Base and superstructure
Haldon’s argument claims to be making an effort to resolve ‘accusations, sometimes reasonably grounded, of economic essentialism or reductionism’ (p.213). Yet in trying to find a middle ground between the ‘analytical Marxism’ of John Roemer and John Elster on the one hand, and ‘the structuralism of thinkers like Althusser’ (p.205), he has not found any dialectical core of Marxism. In a discussion of base and superstructure in the Byzantine Empire, Haldon argues that belief, Christianity in this case, needs to be worked into an analytical framework. While religion ‘was not in itself a relation of production in any way’, it nonetheless impacted upon the economic base (pp.229-30).
This analysis seems not to have overcome any of the limitations of the reductive understanding of base and superstructure that it was attempting to accomplish. There is in fact a reductive approach in the very formulation of the terms of debate about ‘religion’. Theology certainly belongs in the ‘superstructural’ category, but the different Christian churches were very material institutions, which embedded and enforced particular social relations in their host societies. The medieval split between Catholic Christianity and Orthodoxy was nominally over quite obscure theological points, but reflected very real chasms between the social organisation prevalent in the two successor halves of the Roman Empire.
The two social institutions were driven to find points of theological difference, because the logic of their practice was embedded in two very different polities. One existed without a functioning secular state, and the other with a very rigid and dominant one. The result was that the two Churches could not co-operate. This is witnessed in the largely illiterate, Catholic West’s total inability to grasp the nature of the iconoclast controversies in the rather more literate East. Misunderstandings over these kinds of issues, at the level of social consciousness, made a split unavoidable.
Social being determines social consciousness, but social being needs to be understood in a relational, dialectical fashion. A Marxist analysis must relate the parts dialectically to a totality, and cannot be served by reductive analytical categories that are defined by fixed internal characteristics, rather than by their dynamic interrelations. The best contributions in this fascinating collection are alive to the way that Marx developed his dialectical approach to analysis, and in a range of detailed case studies, are able to show the fruitful potential the study of modes of production can still provide. Indeed, if Marxists can overcome subservience to positivistic analytical presumptions, and recapture Marx’s own way of thinking, then we are only at the beginning of a potential transformation in our understanding of human history.
 See Jairus Banaji, Theory as History: Essays on Modes of Production and Exploitation (Leiden: Brill 2010), p.50.