Jairus Banaji has been a prolific and brilliant writer on economic historical topics for almost 40 years now, so a publication of some of his core articles and essays seemed long overdue. "Theory as History", first published in Historical Materialism's excellent book series, fills this gap admirably. Focusing on Banaji's specialisms, agricultural history - especially of late antiquity and the early middle ages - as well as historiography, the essays span the entire period of his work at Oxford, SOAS, and as a Platform Left activist in India. The core subject of this book is the Marxist concept of the mode of production, and the difficult and important question of how to delineate different historical modes of production and what empirical evidence can be marshalled to form criteria to settle such a question.
As with all science, Marxist categories are useful if and only if they actually aid the understanding and reduce the confusion from a multiplicity of real forms to a number of basic organizing categories. If however they are imposed on the matter in an attempt to make empirical evidence fit preconceived categories, chances of error increase considerably. This is what Banaji accuses a long tradition of Marxist history-writing of doing, to a greater or lesser degree. And here one has to agree. Not to say that everything produced hitherto is to be thrown out, obviously; but what has been systematically lacking is the kind of in-depth empirical analysis of economic phenomena and forms of production and exchange which is then *followed*, not preceded, by analysing the logic operating in these phenomena and the formulation of a mode of production concept that can usefully categorize that logic as distinguished from other socio-economic logics. It is the use of Marx's notion of the capitalist mode of production that has always dominated thinking on modes of production, and has overshadowed the long prehistory of pre-capitalist modes. Because of the power of Marx's analysis of capitalism in its pure form, Marxist economic historians have tended to want to find a model of the pure form of other modes of production in the past, and wanted to determine this on the basis of 'purely economic' phenomena, just like Marx did with capitalism. The result has been, as Banaji rightly points out in the opening essay, an often rigid and unhelpful formalism as well as an unjustified neglect of major concepts important historically. These latter concepts often play a central role in the thought of historians of other subdisciplines or traditions: phenomena such as conquest, demographic change, mixed forms of labour, religion, and so on. Sadly, in most of the book Banaji does not go into these categories much further, preferring to focus on the question of modes of production, but his work points here to a wider problem in Marxist historiography.
To briefly summarize the brunt of "Theory as History", Banaji's essays make the argument that Marxists hitherto have tended to conflate three different economic categories or levels of analysis: the mode of production, the form of exploitation, and the relations of production (which seem roughly synonymous with the 'organisation of labour'). In particular the former two are important, because it has led many historians wrongly to dismiss or try to ignore evidence for the enormous historical differentiation of forms of exploitation of labour, in the Marxist quest to maintain a set of easily understandable and useful sequences of modes of production. With a lucid and compelling, if somewhat overwhelming, use of historical evidence throughout many different historical places and times, Jairus Banaji demonstrates that a mode of production does not necessarily correspond with a particular form of exploitation, and that therefore the presence of such a form does not of itself tell you what the mode of production is. Examples here are slavery in the Caribbean and the southern United States, which despite their non-capitalist form of exploitation, were nonetheless capitalist production; the persistence of slavery and various forms of bonded, non-serf labor in the early Middle Ages, which thereby does not demonstrate that a 'slave mode of production' persisted, nor that feudalism existed in ancient Rome; the essentially capitalist form of many forms of rural debt peonage and indebted labor, where the 'debt' in reality functions as a wage equivalent and the surplus appears as 'interest', but is really profit - a case with not fully capitalist forms of exploitation formally subsumed under capitalist production, and so forth. Many such cases are discussed, and however much sometimes in an excess of historical detail and minutiae, the point is well taken.
Another major point, immediately connected with the former, is Banaji's efforts to establish the prevalence of various forms of wage-labour throughout history, as well as the prevalence of what he considers to be capitalist forms of exploitation in pre-capitalist modes of production. The former is increasingly recognized in the Marxist historical literature, somewhat belatedly perhaps in comparison with many non-Marxist economic historians (a point the author also notes); the latter is considerably more controversial. He very rightly points to the very developed forms of mercantile capital existing in the late medieval 'Islamic world', in Sung China, and in the trading states like Venice and Genoa. He also makes some suggestive arguments about the significance of mercantile capital for organizing production, not just in the kind of putting-out systems and setting up manufactories, but also in formally subsuming very disparate producers in India and Africa to their logic. Here some more decisive empirical evidence might be needed.
More importantly, however, this also shows some of the lacunae of Banaji's work in this book. In distinguishing capitalist forms of exploitation, especially those by mercantile capital, from the dominance of a capitalist mode of production, he makes 'capitalist' as a description of a society mean some very different things. Given the only possible use and justification of the categories of modes of production can be to clarify the organizing logic fundamental to reproducing a given society, it seems unwise to allow such confusion to appear. Therefore, contrary to Banaji (and to Marx in some places), I would not suggest using 'capitalist' in an undifferentiated sense for the activities of such figures as Roman estate managers or Venetian long-distance traders. I have also long considered the possibility of conceptualizing a separate mode of production to describe the societies dominated by mercantile capital only, roughly from 1300s (in northern Italy etc) to 1800 or so; perhaps a 'mercantilist mode of production' or somesuch. Whatever the case may be, here appears a weakness in the book: while there is much talk of how not to determine modes of production, Banaji says nothing about what he does think is the sequence of modes of production, or what criteria can be used. There is no concern about the organizing logic of reproduction of an entire society as the deciding criterion for what form of exploitation 'predominates', as Marx says - in fact, reproduction is a term not used at all for this analysis in the book. This seems somewhat strange, and opens the author up to accusations of a certain empiricist superficiality. Also, with the focus on agriculture, a theorization of urban manufacturing, the role of urbanization as in the Brenner thesis, and so forth is absent; but of course one can't expect everything in one work.
There are some other problems with the book as well. One is that the final essay, his 'synthesis' on modes of production, is really illuminating for his whole procedure throughout the book's articles - it would have been much easier on the reader's understanding to put this in the front. A similar annoyance is the decision to not publish the different contributions chronologically, so that articles from the 1990s and 2000s are put before articles from the 1970s, at times. This makes some difference as in the earlier articles, Banaji seems more inclined to maintain his clear distinction between the variety of forms of exploitation and the more orthodox Marxist use of modes of production, whereas in later ones, the latter almost disappears from view, the focus being more and more on the specifics of organisation of labour and the forms of exploitation only. Such developments, however, are obscured by the sequence of essays, without this serving a clear thematic purpose. That said, this book is a brilliant contribution to Marxist historiography as well as a great collection of often astonishingly erudite contributions to (Marxist) economic history, and is well worth reading by anyone interested in historiography, agricultural history, or Marxism, or any combination of these. His plea for a more 'thick' Marxist history-writing is one that can only be applauded, and will hopefully lead to a greater convergence between Marxist approaches and those of other historians.