‘Marxism and the Concept of a Social Formation: An Immanent Critique of the Views of

‘In all forms of society ( Gesellschaftsformen ) there is one specific mode of production which predominates over the rest, whose relations thus assign rank and influence to the others. It is a general illumination which bathes all the other colours and modifies their particularity. It is a particular ether which determines the specific gravity of every being which has materialized within it.’


Introduction
Jairus Banaji has made a major contribution academic life in a variety of areas in a career of over five decades.This includes his work as a historian of the ancient world (Banaji, 2001;2011;2016); his work on the role of the peasantry in agrarian societies (whether precapitalist or not), including feudal society in medieval Europe (Banaji, 1976;1990;2002;2009), as well as that of twentieth century India (Banaji, 1972;1973;1975;1977); his recent work on the history of commercial capitalism (Banaji, 2020); and his work on Marxist theory, especially as it has an application to the research practice of historians (Banaji, 1972;1973a-b;1977b;2010).Banaji was a deserved winner of the Isaac Deutscher Memorial Prize in 2010, for his Theory as History: Essays on Modes of Production and Exploitation, a selection of his writings devoted to theme of the interface between history and Marxist theory.The journal Historical Materialism published a symposium largely devoted to a celebration of his work in 2013. 3However, not surprisingly, there has also been a more critical engagement with his writings over the years (e.g.Brass, 2012;2020).My intention in this article is a limited one.I focus on Banaji's contribution to Marxist theory, especially on what he has to say about the concept of a social formation.In Theory as History Banaji is critical of this concept and of those Marxists who attach importance to it.I subject his views on this subject to an immanent critique.The notion of immanent critique might be understood in different ways (Adorno, 1988(Adorno, (1967)) ;Antonio, 1981;Buchwalter, 1991;Helmling, 2005).Here I have in mind simply the idea that the conclusion of an author's argument can be shown to be inconsistent with that argument's premises.Alternatively, we may say that an immanent critique demonstrates that, given an argument's starting assumptions, a different (perhaps even the opposite) conclusion from that which is drawn by its author is possible.Understood in this way, an immanent critique addresses the issue of the consistency of an argument.Such a critique need not involve endorsing the truth of its premises.Applying this view of immanent critique to the work of Banaji, my aim is to demonstrate that his rejection of the concept of a social formation is not well-founded and that an endorsement of the concept would be consistent with his own theoretical assumptions.

On the Concept of a Social Formation
The term 'social formation' (Gesellschaftsformen) is used to refer to what people using non-technical language call a type of society.Part of the point of using it is to express the idea that the 'object' of historical knowledge is a particular society at a particular time, such as English society in the seventeenth century, French society in the eighteenth century, or Russian society in the nineteenth century.However, those Marxists who employ the concept think about all societies in a very particular way.This differs significantly from that which has often (wrongly) been associated with Marxism.The principal source for that view of society is the Preface to Marx's A Contribution to a Critique of Political Economy (Marx, 2010b(Marx, (1859)), or rather a superficial reading of it.
The concept of a social formation is usually associated with structuralist Marxism, especially the work of Louis Althusser.Indeed, it is sometimes (in my view wrongly) identified with it.Use of the concept had some vogue amongst Marxists writing in the nineteen seventies and eighties, largely as a consequence of the influence of Althusser and his ideas (Hindess and Hirst, 1975;1977).Writing in 1980 Perry Anderson claimed that the distinction between the concept of a mode of production and that of a social formation had at that time 'passed into general usage' (Anderson, 1980, 67).This may well have been the case three or four decades ago.However, although it has not fallen entirely into disuse (da Graca and Zingarelli, 2015), the concept of a social formation receives far less attention today than it has done in the past.There are a number of reasons for this decline in interest, one of which is the fact that structuralist philosophy no longer has the influence that it once had amongst Marxist intellectuals.The concept has been rejected by some Marxists because of what are considered to be its undesirable associations with structuralist philosophy in late twentieth century France, the epistemological assumptions of which, as set out by Althusser, are (rightly) considered to be incompatible with those of Marxism properly understood.
The attribution of the concept of a social formation to Louis Althusser is made by Perry Anderson (1980, 67-68).Anderson acknowledges that 'the term "social formation"' is 'taken from the 1859 Introduction' to Marx's Grundrisse, specifically from what I refer to above as the focal passage (Anderson, 1980, 67;Marx, 1973Marx, (1859), 107)), 107).However, despite this, he maintains that it was Louis Althusser who, in Reading Capital, 'invented' the 'distinction between mode of production and social formation,' and that 'the notion of social formation itself had little or no currency within Marxism prior to Althusser' (Anderson, 1980, 67).This view seems to me to me to be mistaken, for two reasons.In the first place, the concept (understood in its most comprehensive sense) is more closely associated with the work of Etienne Balibar than with that of Louis Althusser.In the second place, as Anderson concedes, it can be found in Marx's own writings.Marx alludes to this concept (without giving it a name, or associating it with a technical theoretical term) in the focal passage from the Grundrisse cited above.However, he also refers to it elsewhere.Indeed, he deploys it to good effect when discussing the transition from feudalism to capitalism.
According to Balibar, the justification for using the term 'social formation' ('formation sociale') is to address two main weaknesses of a certain way of thinking about Marx and Marxism.These both emerge as a consequence, not so much of an excessive reliance on Marx's Preface but, rather, on a superficial reading of it, which is usually associated with the notion of orthodox or vulgar Marxism by its critics, including Jairus Banaji.As both Perry Anderson and David Harvey have noted, Balibar's conception of society as a social formation departs from the view which is thought to be expressed by Marx in the Preface, in two key respects (Anderson, 1980, 68;Harvey, 2006, 25-26).
Firstly, a society understood as a social formation is associated with more than one mode of production.As Balibar has put it, some Marxists have a tendency to assume (wrongly) that a particular society may be 'related to (just) one mode of production' (Balibar, 2015(Balibar, (1968), 365-66fn5)), 365-66fn5).Balibar states that a society understood as a social formation is composed of 'a totality of instances articulated on the basis of a determinate mode of production' (Balibar, 2015(Balibar, (1968), 365-66fn5)), 365-66fn5).What Balibar has in mind here when he talks about 'instances' and about their being 'articulated' need not detain us.The crucial point, once the structuralist phraseology has been stripped away, is his rather common sense but nonetheless significant claim that Marxists should accept that the economic base or foundation of any particular society may contain, and indeed usually does contain, at least two modes of production in combination and interacting with one another.I shall refer to this as the idea of 'modal combination.' The second way in which Balibar's ideas differ from those which are sometimes associated with Marx's Preface, and a further reason for thinking that the concept of a social formation is a valuable theoretical tool for Marxist historians, is because those who employ it reject the idea of economic, technological or productive-forces determinism.Historians of this kind do not, or need not, entirely abandon the 'base-superstructure' distinction that is used by Marx in the Preface.However, contrary to the view which is so often wrongly associated with Marxism, they think in terms of a much looser relationship between these two spheres of society, which relies on the notion of 'conditioning' rather than 'determination' ('Bestimmen') (Williams, 1977, 83-84); and on that of the reciprocal interaction between a society's economic base and its legal, political and ideological superstructure.I shall refer to this as the notion of 'reciprocal interaction.'Those Marxist historians who think that the concept of a social formation is a valuable theoretical tool do so in part because they think it helps them to grasp the complexities of their subject matter when analysing concrete historical examples.For example, Perry Anderson considers the explicit introduction of the concept of a social formation to the Marxist lexicon by Althusser (sic) in the nineteen sixties to be an 'historiographic advance' which 'demonstrably permits' and leads towards 'greater discrimination and complexity in the investigation of concrete societies ' (Anderson, 1980, 68).It is for this reason that he utilised the concept in his own historical writings in the nineteen seventies.
Anderson maintains in his Passages from Antiquity to Feudalism that the point of using the concept of a social formation is 'to underline the plurality and heterogeneity of possible modes of production within any given historical and social totality.'He claims there that 'every concrete social formation is always a specific combination of different modes of production, and those of Antiquity were no exception.'Social formations, he continues, are 'always concrete combinations of different modes of production, organized under the dominance of one of them.' Anderson thinks that employment of the concept of a social formation, especially but not only because of its association with the idea of modal combination, which enables Marxist historians to engage in more sophisticated historical investigations than is possible for those who rely on a superficial reading of the text of Marx's Preface to A Contribution to a Critique of Political Economy.

Banaji's Critique of Vulgar Marxism
Bearing the above in mind, let us now turn to consider the views of Jairus Banaji, as expressed especially but not only in his Theory as History: Essays on Modes of Production and Exploitation (2010).Banaji is critical of those Marxists who employ the concept of a social formation.He associates the concept with what he refers to as 'vulgar Marxism' (Banaji, 2010, 46-49, 53, 61-65, 150).He characterises vulgar Marxism as an 'ossified pseudo-Marxism' (Banaji, 1977, 11).By this he has in mind, not the classical Marxism of Marx, Engels, Lenin and Trotsky, but rather the Marxism of Karl Kautsky, Georg Plekhanov and the Second International, together with that which is associated with Stalin and Stalinism, which he understands in a certain way (Banaji, 2010, 49).Banaji maintains that this kind of Marxism is both dogmatic and doctrinaire.It is quasireligious or 'scholastic' in its demands for at least lip-service to be granted towards a given theoretical orthodoxy which, in his view, is insensitive to the specificities of history and 'has little to do with Marx himself' (Banaji, 2014, 2-3;Banaji, 1977a, 11;Banaji, 2010, 49;Banaji, 2013, 131).This assessment of the ideas of Karl Kautsky, as the defender of the principles of orthodox Marxism, is not uncommon.Melvin Rader refers at one point to 'the mechanistic determinism of Kautsky, Plekhanov, and other theorists of the Second International,' all of whom he associates with 'fundamentalist' Marxism (Rader, 2979, 5fn4).Chris Harman has also associated Kautsky with deterministic 'mechanical' materialism, describing him as 'the "Pope of Marxism" (Harman, 1998a, 9-10, 34, 51).Similarly, Raphael Samuel, has characterised Kautsky as 'the high priest of orthodox Marxism in the epoch of the Second International' (Samuel, 1980, 25).On the other hand, others have claimed that this assessment of Kautsky and his ideas, in particular, is based on a highly selective reading of his writings (Townshend, 1984;Townshend, 1989; see also Burns, 2000a, 15-17;Blackledge, 2005, 53, 56-59;Blackledge, 2006).
It is fruitful at this point for us to distinguish what Banaji refers to as vulgar Marxism, and others as orthodox Marxism, from what I propose to call classical Marxism, or the ideas of Marx and later Marxists properly understood.It is the former and not the latter which is associated with mechanistic determinism and reductionism.Banaji might perhaps be criticised for 'buying-in,' at times, to this reading of Marx's views on history.It is true that his intention is to criticise it.Nevertheless it is arguable that, in so doing, he might perhaps be said at the same time (paradoxically) to presuppose it.Those who read the ideas of Kautsky in a non-deterministic way would regard Banaji's critique of vulgar Marxism as misplaced, precisely because of the erroneous assumptions upon which it is based.In their view there is an element of a 'straw person' about the target of Banaji's criticism. 4  So far as history and historical writing are concerned, Banaji makes three criticisms of vulgar Marxists, which pull in different directions.The first of these criticisms is that they ignore history altogether and advance a form of Marxism which is purely philosophical or theoretical.To invoke the colour imagery that is employed by Goethe in Faust Part One, in the quotation which stands at the head of this paper, the outlook of philosophy (or perhaps of one way of thinking about it), which stands opposed to that of history, is profoundly 'anti-life.'The colour of philosophy is grey.On this view, as Hegel suggests in the Preface to his Philosophy of Right, philosophy necessarily 'paints its grey in grey' (Hegel, 1979(Hegel, (1821), 13)), 13).The outlook of history and of the historian, on the other hand, being life-affirming, is best represented by the colour green, for 'green is the golden tree of life' (sic).Needless to say, of course, for Hegel the method of philosophy, properly understood, incorporates both of these contrasting insights.Hegel suggests that the true philosopher is always and necessarily also a good historian, although whether or not Hegel himself should be regarded as a good historian is another matter.
Banaji's second criticism of vulgar Marxists is that they have a tendency to talk about historical method, or about how history should be written, without actually writing any history themselves.Banaji has very little time for those who do this.When making this criticism he appears to have especially in mind those who are associated with structuralist Marxism, including Louis Althusser and his followers in France, as well as Barry Hindess and Paul Hirst in England (Althusser, Balibar, Establet, Macherey andRancière, 2015 (1965); Hindess and Hirst, 1975;Hindess and Hirst, 1977;Hirst, 1985).Rightly or wrongly, Banaji considers structuralist Marxism to be a specific form of vulgar Marxism.He is therefore critical of it (Banaji, 19772010, 8fn28, 184fn16, 212;2011, 112fn16).For example, at one point, he maintains that those structuralist Marxists who employ the concept of a social formation embrace a 'motionless paradigm' (Banaji, 2010, 359).To employ the technical vocabulary of structuralist philosophy, their approach is entirely synchronic and not at all diachronic.As such, it has to do more with 'archaeology,' in the sense in which Michel Foucault employs the term in The Archaeology of Knowledge, than with history, which necessarily has to do with processes of change over time (Foucault, 2002(Foucault, (1968))).On another occasion, Banaji observes that structuralist Marxists such as Barry Hindess and Paul Hirst are 'generally quite ignorant of real historical processes' (Banaji, 1977.Nevertheless, despite the specific objections which he has to structuralist Marxism, Banaji's main target is vulgar Marxism in general.It is vulgar Marxists tout court who employ the concept of a social formation and who are open to criticism because, in his opinion, their approach does not attach sufficient importance to the complexities of history.
Banaji's third criticism is that, if and when they do attempt to write about history, vulgar Marxists employ not an empirical but rather an a priori method.This involves reading into historical situations things which are not in fact discovered by empirical investigation.They make assumptions in advance about what must be present in the situation they are examining, given the prior commitment which they have to a Marxist theoretical framework.Moreover, they consider the application of that framework to the situation in question to be valid independently of the factual evidence that is required for their conclusions to be justified.
Developing this third criticism further, Banaji argues in Theory as History that vulgar Marxists have a fondness for 'scholastic disquisitions' and for 'abstract historical formalism' (Banaji, 2010, 49).On more than one occasion he characterises vulgar Marxism as purely formal or formalist, that is to say, formulaic (Banaji, 1977a, 3-4, 6, 11;Banaji, 1977bBanaji, , 1390Banaji, , 1398Banaji, , 1400;;Banaji, 1979b, 486-87;Banaji, 2010, 8, 49, 61, 308, 346, 351;Banaji, 2013, 131, 135;Banaji, 2014, 4).For example, he maintains at one point that they do nothing more than regurgitate abstract 'dead formulas' (Banaji, 1979b, 486), which have little connection with any empirically observable reality.Elsewhere he associates vulgar Marxism with 'formalist a priorism' (Banaji, 1977b(Banaji, , 1400)).In short, vulgar Marxists are not sufficiently sensitive to the complexities of the historical processes and situations about which they claim to be writing.John Haldon has rightly pointed out that Banaji's main target, generally speaking, is those Marxists who do not pay sufficient attention to 'historical complexity' (Banaji, 2010, 22;Haldon, 2013).A pre-requisite for this is the empirical investigation of concrete situations which in Banaji's opinion necessarily lies at the heart, not only of all good history writing, but of any and all writing that might properly be described as 'historical.'This is an essential component element of the craft of the historian.
As Banaji points out, his critique of the formalism or excessive theoreticism of vulgar Marxism echoes concerns which Marx himself expressed about the way in which Russian Marxists such as N. K. Mikhailovsky understood his views in the nineteenth century (Banaji, 2010, 2, 351).In a letter written in November 1877, which he sent to the editor of the Russian journal Notes on the Fatherland (Otechestvenniye Zapisky), Marx takes issue with Mikhailovsky's understanding of his views on history (Marx, 2010a(Marx, (1877), 196-201)), 196-201).He argues forcefully that Mikhailovsky has transformed the 'historical sketch of the genesis of capitalism in Western Europe' which he sets out in volume one of Capital 'into an historico-philosophic theory of the marche generale [general path] imposed by fate upon every people, whatever the historic circumstances in which it finds itself' (Marx, 2010a(Marx, (1877), 200)), 200).Rather than regard the assumptions of the materialist conception of history as a useful method to be applied by historians in their empirical research, Mikhailovsky has transformed those assumptions into a 'universal passport (passepartout) of a general historico-philosophical theory, the supreme virtue of which consists in being super-historical' (Marx, 2010a(Marx, (1877), 201)), 201).
Banaji maintains that the concept of a social formation is employed by those who, although they profess to be Marxists, have forgotten or overlooked the fact that Marx was amongst other things (and perhaps above all) a historian, that is to say somebody who held that, although historical explanation does need to be theoretically informed, nevertheless at the same time it must always rest on concrete analysis and detailed empirical investigation.In contrast to the writings of vulgar Marxists, past and present, Banaji praises the Annales School of French historians for being much closer to Marx's own approach to history writing, than are those 'scholastic' Marxist theoreticians who employ the concept of a social formation.He argues that the 'strictly professional history' that is associated with this school, although it is 'not known for purely scholastic disquisitions on "modes of production" and "social formations," came far closer to the conceptions of Marx than the whole tradition of abstract historical formalism' which he claims has 'passed for Marxism' and which has 'decisively shaped' discussions of the concept of a mode of production (Banaji, 2010, 49-50).

Banaji's and the Philosophy of Hegel
Banaji has some extremely interesting things to say about the philosophy of Hegel (Banaji, 1979a-b).An appreciation of his understanding of Hegel's philosophy is a necessary precondition for a grasp of his views regarding the debt to Hegel that was owed by Marx when discussing questions of research method, whether in the discipline of political economy or in that of history.So far as Hegel is concerned, Banaji relies heavily (though by no means exclusively) on the view of 'scientific' method which Hegel sets out in the Preface to his Phenomenology of Spirit (Hegel, 1977(Hegel, (1807)), § §14-17, 8-10).
Although this is not the appropriate place to discuss Banaji's reading of Hegel at any length, nevertheless at least one point is worth noting here.It is that Banaji (quite rightly) does not argue that the views of Hegel and Marx are significantly different from one another because Hegel is an arm-chair philosopher or theorist who relies exclusively on an a priori method, whereas Marx's approach is significantly different from that of Hegel because it is genuinely scientific, not least because of Marx's emphasis on the need for detailed empirical investigation.Nor does Banaji accuse Hegel of committing the 'sin' of formalistic a priorism.Indeed, in an essay entitled 'Gunder Frank in Retreat,' which was published in 1979, Banaji refers with approval to Hegel's critique of 'formalism' in the Preface to is Phenomenology of Spirit (Banaji, 1979b, 487).
When criticising 'formalism' in this Preface, Hegel rejects outright what he considers to be an erroneous philosophical method.This method attaches exclusive importance to the principle of universal 'form' and no significance at all to the empirical 'content,' or to the multiplicity of determinate shapes, in and through which that universal form is presented to the knowing subject in the act of experiencing it.In his opinion, this erroneous philosophical method (which Hegel seems to have associated especially with the philosophy of Schelling among others) is partial or one-sided.For example, it focuses exclusively on the essence of things and ignores completely their appearance, and hence also their concrete actuality.As Hegel says in the Preface to his Philosophy of Right, the one universal and divine Idea, upon its 'actualization,' inevitably 'emerges with an infinite wealth of forms, shapes and appearances' (Hegel, 1979(Hegel, (1821)), 10), all of which he claims are overlooked by the advocates of the philosophical formalism he is criticising.
Employing the imagery of colour, as he also does in the Philosophy of Right, with his reference to the outlook of a philosophy which 'paints its grey in grey ' (1979 (1821), 13), Hegel argues in the Preface to the Phenomenology that the 'monochromatic formalism' which he rejects might be associated with the 'undoing of all distinct, determinate entities (or rather the hurling of them all into the abyss of vacuity,' which is, as he famously puts it, 'the night in which, as the saying goes, all cows are black' (Hegel, 1977(Hegel, (1807)), § §15-16, 9).
Banaji is extremely sympathetic towards Hegel's critique of formalism in the Preface to the Phenomenology.It is clear enough, therefore, that (again quite rightly) he does not consider Hegel's philosophy to be itself subject to a critique along these lines.This is perhaps a difference between Banaji and Marx, who sometimes gives his readers the impression that he does (in my view wrongly) associate Hegel's philosophy with formalism in Banaji's sense of the term, that is to say, with excessive rationalism and a methodological approach which is entirely a prioristic (Marx, 1967(Marx, (1844)); Marx, 1973bMarx, (1846), 93;), 93;Marx 1974Marx (1867)), 29; Marx andEngels, 1954 (1844), 78-82).

Banaji on Marx and 'Scientific' Method
Banaji's critique of vulgar Marxism raises interesting questions in relation to Marx's views on 'scientific' research method, especially but not only in the General Introduction to the Grundrisse manuscript.This is a subject which has in the past received quite a lot of discussion, especially in the period immediately following the publication of Martin Nicolaus' English translation of the Grundrisse in 1973 (Carver, 1975;Carver, 1980;Echeverria, 1978;Echeverria, 1980;Mepham, 1978;Tribe, 1974;Sayer, 1981; see also Burns, 2000).It is also considered at length by Banaji, in a paper entitled "From the Commodity to Capital: Hegel's Dialectic in Marx's Capital," which was published in 1979, but which is not reprinted in Theory as History (Banaji, 1979b).
When discussing this issue Banaji strongly emphasises the significance of Marx's debt to the philosophy of Hegel (Banaji, 1979b, 17-21).Banaji rightly observes that there are at least some striking similarities between the views of Hegel and those of Marx regarding the issue of what a properly 'scientific' research method looks like.When discussing Marx's debt to Hegel, he refers to 'the enormous weight of Hegel's method in Marx's development' (Banaji, 1979b, 14).Following Lenin, he insists that 'it is impossible to grasp Marx's conception of scientific method outside the framework of Hegel's Logic.' (Banaji, 1979b, 19).Not surprisingly, this leads him to reject all interpretations of Marx's views on method which seek to downplay the influence of Hegel's philosophy upon him.These include the interpretations of Galvano Della Volpe, Lucio Colletti and Louis Althusser (Banaji, 1979b, 14-15, 25-26, 36, 42-43; see also Burns, 2000).
Of particular interest here are Marx's views regarding the levels of theoretical analysis which are associated with the explanatory concepts that are employed by researchers within the fields of history and political economy.As Banaji points out, Marx follows Hegel by classifying these as being either 'abstract' on the one hand or 'concrete' on the other (Marx, 1973a(Marx, (1857-58), 81-112;-58), 81-112;Banaji, 1979b, 17-21).
Marx suggests, both in the Introduction to the Grundrisse and in the Afterword to the second German edition of volume one of Capital (Marx, 1974(Marx, (1867), 28)), 28), that whether we are talking about questions of political economy or questions of history we should not think in either-or terms about a purely a priori, philosophical or ahistorical conceptual method, on the one hand, and a research method which involves atheoretical, empirical investigation (in the manner of empiricism or positivism) on the other.Rather, he maintains that an appropriate research method requires both historians and students of political economy to steer a middle course between these two extremes.On this issue Marx in fact follows Hegel, although without perhaps appreciating (or acknowledging) that he is doing so.I say this because, as I suggested earlier, Marx sometimes accuses Hegel of formalism, as Banaji understands the term.
Like Hegel properly understood, Marx too was opposed to what Banaji refers to as formalism.He rejected a priorism in the writing of history and therefore also excessive theoreticism.I use the word 'excessive' here because, as John Hoffman has noted, when discussing questions of 'scientific' method Marx acknowledges that the appropriate method for the presentation of results does involve an element of a priori reasoning (Hoffman, 1975, 120).If researchers do their job well then, once their results have been presented, it will, Marx argues, appear as if we have before us a mere a priori construction' (Marx, 1974(Marx, (1867), 28)), 28).Marx nevertheless insists that we must draw a sharp distinction between the appropriate method of presentation of results and 'the method of enquiry,' which he accepts should be 'severely realistic.'That is to say, it necessarily requires some act of empirical investigation, it being understood, as Harvey J. Kaye has observed in a discussion of the ideas of E. P. Thompson, that the terms 'empirical' and 'empiricist' ought not to be identified with one another (Kaye, 1984: 209, 211).Hoffman has rightly said that the fact that Marxist theory is presented in 'an a prioristic manner' does not mean that 'it has not been derived form a most throughgoing investigation of the real world' (Hoffman, 1975, 120).
The need to balance or juxtapose an empirical method of enquiry and what might be termed a 'rationalist' method of presentation of results lies at the very heart of Marx's views regarding scientific method, whether in the discipline of political economy or in that of history.This brings Marx very close to the position advanced by contemporary realist philosophers of science such as Roy Bhaskar, with his emphasis that scientific research leads to the production of propositions which (paradoxical though it may seem) are both 'analytic' truths and yet at the same time also knowable a posteriori (Burns, 2000, 86-90).
Marx's treatment of questions of method, in the Grundrisse and elsewhere, suggests the need for a type of writing which although it is necessarily theoretically informed nevertheless also attaches due importance to the complexity of particular historical situations.This involves paying due attention to factual details, which in turn requires, not armchair theorising, but rather serious empirical investigation, together with a willingness to revise one's own theoretical assumptions in the light of it.This is Banaji's view also, most of the time.At one point he refers favourably to the work of Michael Cowen, sharply contrasting it with the structuralist Marxism of Barry Hindess and Paul Hirst, precisely because of its 'combination of empirical content and theoretical sophistication' (Banaji, 1977Banaji, 2010, 309fn78;Cowen, 1976a-b).
It should however be noted that Banaji does occasionally appear to stray too far ('bend the stick') in the opposite direction.He is so concerned with avoiding the pitfalls of vulgar Marxism in historical writing, especially the 'sins' of theoreticism, formalism and a priorism, that he comes very close at times to embracing the standpoint of empiricism.For example, in an essay entitled 'André Gunder Frank in Retreat,' published in 1979, having emphasised the need for Marxists historians to carry out research on 'the forms in which wage-labour is created, on the one hand, and the forms in which it is initially subsumed into capital, on the other,' Banaji goes on to claim that in this particular area of study 'there are no models to guide us.' Indeed, he argues, this involves an act of empirical research which must be 'entirely non-aprioristic' (my emphasis), precisely because and insofar as 'it is historical' (Banaji, 1979b, 485).This seems to me to take a good idea too far.Nor is it consistent with what Banaji says about this issue elsewhere, when he insists that the empirical investigations of the scientist are always and necessarily theoretically informed.

Modes of Production and Mechanisms of Exploitation
We may illustrate what Banaji has in mind when he criticises vulgar Marxists and their employment of the concept of a social formation by considering what he has to say about modes of production and the mechanisms of exploitation with which they are associated.Banaji's main reason for objecting to the use of the concept of a social formation is because he believes that the employment of this terminology is necessarily associated with two things of which he disapproves.First it is associated with the notion of the articulation of different modes of production.Second it is associated with the view that each mode of production in such an articulation must be associated with just one mechanism of exploitation.Banaji argues that Maurice Dobb, Ernesto Laclau and Chris Wickham are all guilty of making this second assumption, which he considers to be erroneous (Banaji, 2010, 25, 41, 52-55, 58, 61, 68, 213, 347, 359).
In Theory as History Banaji insists that 'relations of production are simply not reducible to forms of exploitation,' not only because 'modes of production embrace a wider range of relationships than those in their immediate process of production,' but also because 'the organisation and control of the labour-process, 'correlates' with historical relations of production in complex ways' (Banaji, 2010, 41).So far as the capitalist mode of production is concerned, Banaji maintains that there is a very significant difference between the situation of 'the potters of Moscow' and that of the slave owners and their slaves in 'the slave South' of the United States, or 'the sugar latifundia of coastal Peru.'He argues that 'in all these varied instances' of capitalist exploitation there is 'no question of identifying the "mode of production" according to the character of the given forms or relations of exploitation.'Nor, in his view, 'did any of these instances involve a "coexistence" of modes of production' (Banaji, 2010, 58).Banaji concludes that 'capitalist relations of production are compatible with a wide variety of forms of labour' and of exploitation, 'from chattel-slavery, sharecropping, or the domination of casual labour markets, to the coerced wage-labour peculiar to colonial regimes and, of course, "free" wage-labour' (Banaji, 2010, 359).
Similarly, when discussing the feudal mode of production in Western Europe in the medieval period, Banaji argues that 'the feudal enterprise was sustained by a variety of forms of labour, comprising domestic servants who were legally slaves and who undertook the principal tasks, especially ploughing; day-labourers who housed separately on the estate; part-time hired workers recruited from impoverished peasantry, free tenants who performed seasonal or supplementary services; and the serf-population as normally understood, i.e. villeins bound by labour-services' (Banaji, 2010, 92).
If we consider the situation of 'slaves' and of 'hired labourers' (wage employees) in the economic systems of medieval society in Western Europe then those Marxists who deploy the concept of a social formation would argue that these different forms of exploitation, although they were undoubtedly present in feudal social formations, should not be thought of as being component elements of the feudal mode of production, in the strict sense of the term.Rather, they are component elements of a feudal social formation.This social formation is designated as 'feudal,' not because other modes of production were absent from it, but, rather, because the feudal mode of production was dominant within it.However, Banaji denies all of this.As he puts it, 'the slaves and hired labourers who intervened in this type of economy were as much part of specifically feudal relations production as the serf-population itself' (Banaji, 2010, 92).Consequently, their presence in medieval society 'did not signify the persistence or emergence of other relations of production.' Nor, therefore, does it 'imply an "articulation" of several distinct "modes production",' as is claimed by those vulgar Marxists who employ the concept of a social formation (Banaji, 2010, 92).
According to Banaji, modes of production and their associated relationships of exploitation can be concretely manifested in different ways.Marxist historians need to be sensitive to this fact.If they are at all interested in grasping the complex circumstances of particular historical situations, as they should be, then this requires that they seriously consider the different 'possible ways in which a mode production can be configured historically' (Banaji, 2010, 22-23).Excessive reliance on 'abstract concepts,' he argues, is likely to lead the historian to blur over, or attach insufficient importance to, these observable empirical differences.In his view, the employment by Marxists of the concept of a social formation is likely to, and often does in fact, lead to a failure on their part to grasp the complexity of the historical situations or processes about which they are writing.In Banaji's own words, the employment by Marxists of the concept of a social formation is likely to 'obscure' the specificities of the various forms of exploitation which can be and are associated with a particular mode of production, in its different possible concrete manifestations.It is, therefore, symptomatic of a descent into 'vulgarity.'As Banaji puts it, 'although it is in some sense quite self-evident and banal' (sic), nevertheless, 'the distinction between "modes of production" and "social formations" that is generally drawn in most recent Marxist literature may actually obscure and mystify' these different mechanisms of exploitation (Banaji, 2010, 92).
None of this leads Banaji to argue that Marxists should stop talking about modes of production.However, it does lead him to the conclusion that they should stop talking about any given mode of production as if in principle it could only be associated with just one mechanism of labour exploitation, as he claims is assumed by those vulgar Marxists who employ the concept of a social formation.One implication of this view is that, given the variety of mechanisms of exploitation which are associated with any given mode of production, it follows that in Banaji's opinion it is not actually necessary for Marxist historians to make the distinction between the concept of a mode of production and that of a social formation.Banaji appreciates that, according to those Marxist historians who employ the concept of a social formation, one of the reasons for including it within the lexicon of Marxism is to allow Marxist historians to grasp the fact that in any given society at any given time there will almost certainly, as a matter of fact, be more than one mechanism of exploitation present.In his view, however, the theoretical work that is done by deploying this concept can be and should be done in a different way, simply by changing how Marxists think about the concept of a particular mode of production and the various different forms of relations of exploitation which might be associated with it.

A Reply to Banaji's Critique of the Concept of a Social Formation
It is of course possible, as Banaji suggests, to overlook completely the diversity of concrete forms of exploitation which are associated with the capitalist mode of production in a particular society at a particular time.This is the mistake which, according to Banaji, is made by vulgar Marxists generally.However, it is also possible to make the equally significant mistake of attaching too much importance to that empirically observable diversity.As Lenin argues in The Development of Capitalism in Russia, we might commit the opposite error of 'not seeing the wood for the trees' (Lenin, 2009(Lenin, (1899), 455)), 455).It is, therefore, worth looking at what Lenin has to say about this issue.
In Lenin's opinion, the concept of a social formation is of fundamental significance for Marx.Indeed, he maintains that Marx's views on society and on history in the Preface to A Contribution to a Critique of Political Economy, may be summarised by 'the single fundamental concept: social formation' (Lenin, 2008a(Lenin, (1894), 140)), 140).The concept of a social formation, in the sense indicated at the beginning of this article, as involving both the notion of modal combination, as well as that of reciprocal interaction, can clearly be found in Lenin's writings.When commenting on Marx's ideas, Lenin frequently employs the expression 'social formation.'This term is often employed in the English translation of the Collected Works, as a rendering of the Russian 'obshchestvennaya formatsiya' (Lenin, 2008a (1894), 140-41, 144, 146, 155, 158, 162, 165, 179fn, 182, 189-90; Lenin, 2008b  (1894), 350, 413, 444). 5  The idea that the concept of a social formation might be connected to that of reciprocal interaction is emphasised in Lenin's What the "Friends of the People" Are and How They Fight the Social Democrats (Lenin, 2008a(Lenin, , (1894))).However, so far as the views of Jairus Banaji are concerned, it is what Lenin has to say about modal combination which is most significant.I shall, therefore, not consider further what Lenin has to say about the idea of reciprocal interaction.The association which the concept of a social formation has with the idea of modal combination is addressed in Lenin's The Development of Capitalism in Russia (Lenin, 2009(Lenin, , (1899))).Let us therefore turn to consider what Lenin has to say about the concept of a social formation and its relation to that of modal combination in that text.
When discussing the historical development of Russian society in the nineteenth century, Lenin maintains that Marxist historians need to focus on questions of political economy, specifically the emergence and development of capitalism in this period.He argues that this involved the emergence and later expansion of the capitalist mode of production, not just in the cities, but also in the countryside, in the agrarian sector of the economy.More specifically, it involved the replacement or substitution of what Lenin refers to as the 'Corvée economy,' associated with the feudal mode of production, by capitalist relationships of production as understood by Marx in volume one of Capital.It is for this reason that the economic system of Russian society, and indeed Russian society itself, at this time should be thought of as being in a period of transition.As Lenin puts it, 'capitalist economy could not emerge at once, and Corvée economy could not disappear at once' within the Russian social formation.'The only possible system of economy' was, therefore, 'a transitional one,' that is to say, an economic 'system combining the features of both the Corvée and the capitalist systems' of production (Lenin, 2008(Lenin, (1899), 194)), 194).
According to Lenin, then, the key to understanding the history of Russia in the nineteenth century is to identify the presence there of just two modes of production in combination with one another.Lenin refers in this connection to a 'whole variety of forms of contemporary landlord farming' which, nevertheless, may be in essence be regarded as amounting to just 'two systems,' namely, 'the labour-service and the capitalist systems, in various combinations' with one another (Lenin, 2008(Lenin, (1899), 198)), 198).He insists on the fact that despite 'all the endless variety of forms characteristic of a transitional epoch,' nevertheless, in the final analysis, 'the economic organisation of contemporary landlord farming' in nineteenth century Russia, amounted to the presence of just 'two main systems,' albeit again 'in the most varied combinations' with one another, which he refers to as 'the labour-service system and the capitalist system' respectively (Lenin, 2008(Lenin, (1899), 194)), 194).In The Development of Capitalism, therefore, Lenin evidently does possess the notion of modal combination, which is central to the concept of a social formation as Marx understood it.
Lenin's view that the economic system of the Russian social formation in the nineteenth century contained two modes of production in combination with one another does not necessarily imply a rejection of determinism and reductionism on his part.The result of employing this terminology could in principle be a more sophisticated version of that doctrine.However, there is at least some evidence which supports Etienne Balibar's view that in Lenin's thought the notion of a social formation is associated with a nondeterministic understanding of the materialist conception of history, which attaches due importance to the significance for historical explanation of the relatively autonomous workings of the superstructural phenomena referred to in Marx's Preface to A Contribution to a Critique of Political Economy.
For those classical Marxists like Lenin who employ the concept of a social formation the problems relating to its use which concern Jairus Banaji do not arise (Lenin, 2008a(Lenin, (1894), 140-41, 144, 146, 155, 158, 162, 165, 179fn, 182, 189-90;), 140-41, 144, 146, 155, 158, 162, 165, 179fn, 182, 189-90;Lenin, 2008bLenin, (1894), 350, 413, 444)), 350, 413, 444).This is clear from Lenin's treatment of the Russian example.Lenin insists that we can talk about the presence of just two modes of production in nineteenth century Russia in combination with one another, even though he concedes that there are more than two concrete forms of labour exploitation which are present.This view does not involve a contradiction because, for Lenin as for Marx, and indeed for Jairus Banaji also, theoretical analysis may take place at different levels of abstraction.In the case of the capitalist mode of production, recognition of the existence of different concrete forms of labour exploitation is consistent with the view that, nevertheless, these are all species of the same underlying genus.They are specific instantiations of an underlying essential relation which is associated with the presence of wage-labour, which is necessarily manifested in one concrete form or another.In short, Lenin suggests that labour could in principle be, and in nineteenth century Russia actually was, 'subsumed' under capital in a variety of different ways, to such an extent that the presence of capitalist relations of exploitation may be mystified, or not be readily apparent.This is something which Banaji appears willing to acknowledge on more than one occasion (Banaji, 1977(Banaji, , 1375(Banaji, -76, 1390(Banaji, , 1397(Banaji, -98, 1400;;Banaji, 2010, 277-82, 301-10, 324-32).
Ironically, Banaji points out that one of the sources for his own views regarding this issue is Lenin's The Development of Capitalism in Russia.As Banaji puts it, 'Lenin's pages on the labour-service system in Russia, with their fine distinctions between "bonded hire" and "purely capitalist wage-labour"' are a 'model of how Marxists can restore a sense of complexity to their analysis of exploitation' (Banaji, 2010, 41).Banaji praises Lenin's basic methodological approach.However, when doing so, he overlooks the fact that for Lenin the complexity and diversity he observed through detailed empirical investigation, both can and should be reduced to an underlying simplicity by establishing the existence in nineteenth century Russian society of just two modes of production, in combination with one another.Lenin evidently did not think of this as an over-simplification, but rather as a necessary simplification if we are to be able to grasp theoretically the complexities of historical reality.
In Theory as History Banaji maintains that it would 'represent an advance in Marxist theory to think of capitalism,' as a mode of production, 'working through a multiplicity of forms of exploitation based on wage-labour' (Banaji, 2010, 145).He argues that 'instead of seeing wage-labour as one form of exploitation among many, alongside sharecropping, labour tenancy, and various kinds of bonded labour,' it would be preferable to think that 'these specific individual forms of exploitation may just be ways in which paid labour is recruited, exploited, and controlled by employers' (Banaji, 2010, 145).This implies that in Banaji's view, as in that of Lenin before him, there is something constant, essential and fundamental which underlies this empirically observable variety, namely the wage-relation itself, which may be manifested in a variety of concrete forms, the actual presence or absence of which can only be established by historical research, which always involves detailed empirical investigation.
Banaji claims that his argument 'is not that all [his emphasis] sharecroppers, labourtenants, and bonded labourers' should be regarded as 'wage-workers' (Banaji, 2010, 145).However, it seems to me that this claim is not consistent with his view that these concrete forms of exploitation might in fact 'reflect the subsumption of labour into capital' in different ways, because 'the "sale" of labour-power for wages is mediated and possibly disguised in more complex arrangements' (Banaji, 2010, 145).When making this remark, Banaji acknowledges that within the capitalist mode of production wage-labour should not be seen as just 'one [his emphasis] form of exploitation among many' (Banaji, 2010, 145).Or rather, more accurately, wage-labour should not be regarded as just one concrete from of exploitation among many.In effect, therefore, he agrees with Lenin that wage-labour in the capitalist mode of production has a special status because it is the one and only essential relationship of exploitation for that particular mode, even though it can be and is in fact concretely manifested in a variety of different ways.
The plurality and diversity of the different mechanisms of labour exploitation which are associated with any given mode of production, to which Banaji draws our attention, is not in fact overlooked by those classical Marxists (such as Marx himself and Lenin) who employ the concept of a social formation in their writings.On the contrary, they assume that the point of introducing the concept is to develop a theoretical framework which will make it possible for historians to capture and express this complex variety in theoretical terms.As we have seen in the case of Lenin, the idea of modal combination, or of the articulation modes of production, is central to such a theoretical enterprise.
More recently, a number of commentators have claimed that the concept of a social formation helps us to achieve a better understanding of the complexities of history than is possible if we rely on the over-simplified assumptions of Marx's Preface to A Contribution to a Critique of Political Economy.In their view, if we are to adequately grasp this historical complexity, then, again, we are required to employ the concept of a social formation rather than, as Banaji suggests, to abandon it.As we have seen, this is true of Perry Anderson (Anderson, 1980, 68;Anderson, 1986Anderson, [1974]], 22).However, it is also true of other Marxist historians and sociologists, for example Chris Wickham and Stuart Hall (Wickham, 1984, 7-8;Wickham, 1985, 169, 189;Wickham, 2008, 8;Wickham, 2015, 141, 143, 145;Hall, 1984, 23, 51, 71).Finally, we may draw attention to the work of those Marxist anthropologists who attempted to apply 'articulation theory' to a variety of empirical case studies in the 1970s and 1980s (Harries, 1985;Hopkins, 1978;Scott, 1976;Soiffer, 1982; see also Foster-Carter, 1978).

Conclusion
In this article I have examined the views of Jairus Banaji regarding the significance of the concept of a social formation for Marxism.Banaji maintains in effect that the concept is inextricably associated with 'orthodox Marxism,' or with 'vulgar Marxism,' that is to say with a particular understanding of Marx's views on history that is both reductionist and determinist, and which is based solely on a casual reading of just one text, namely the Preface to Marx's A Contribution to a Critique of Political Economy.Banaji rightly considers this understanding of Marx's views on history to be erroneous.He also thinks (again rightly) that Marx's version of historical materialism, properly understood, is neither reductionist nor determinist.Consequently, he considers it to be necessary to reject the concept of a social formation and to dissociate Marx from the use of it.In my view this argumentative strategy is problematic, simply because the concept of a social formation is deployed by Marx himself, especially when talking about the transition from feudalism to capitalism.It must, therefore, be possible in principle to detach the concept of a social formation from any association which it might be thought to have with orthodox or vulgar Marxism.
At times Banaji associates the concept of a social formation, not merely with vulgar Marxism, but more specifically with the structuralist Marxism of Louis Althusser and his fellow contributors to Reading Capital, including Etienne Balibar.He sympathetic to the line of criticism, often associated with the work of E. P. Thompson, which rejects structuralist Marxism because of its excessive theoreticism (Thompson, 1980, 196;also Bernstein, 2013, 326;Hirst, 1985a;Nield & Seed, 1979).In his opinion, this is another reason for objecting to the concept and for dissociating Marx from its employment.It seems to me, however, that the concept of a social formation is not necessarily associated with structuralist Marxism.On the contrary, it can be found in Marx's own writings.Moreover, when making remarks of this kind Banaji overlooks the very significant differences which exist between orthodox or vulgar Marxism, as he understands it, and structuralist Marxism, not least with respect to their underlying philosophical assumptions, that is to say, their respective views regarding questions of both ontology and epistemology.
There have been numerous occasions when Banaji himself has (directly or indirectly) positively endorsed the notion of a social formation and/or the associated notion of the articulation of a number of different modes of production (Banaji, 1970, 71, 83-84;1973a, 679;1973b, 396;1977a, 23;1977b, 1391;2001, 218-19;2010, 80).For example in an early article, which was in 1970, he refers to the 'Marxist science of social formations' and insists that 'historical materialism alone' offers an adequate 'theory of social formations' (Banaji, 1970, 84).In an article published in 1973, Banaji was happy to cite Nicos Poulantzas's view that 'a social formation is a historically unique object' which, as such, constitutes a '"particular combination or a specific interlocking of several 'pure' modes of production",' for example 'Bismarckian Germany, Stalinist Russia, T'ang China' (Banaji, 1973a, 679;Poulantzas, 2018Poulantzas, (1968), 15)), 15).In an article published in 1977, which was devoted to a discussion of the situation of the small peasantry in the Deccan districts of India, Banaji referred to 'the formal subordination of small producers to monied capitalists' in the Indian 'social formation' in the nineteenth century' (Banaji, 1977b(Banaji, , 1391)).Indeed, even in his 1977 article, 'Modes of Production in a Materialist Conception of History,' which is reprinted in Theory as History and which is generally critical of those who employ the concept, Banaji nevertheless refers himself at one point to 'peasant commodity production' in 'feudal social formations' (Banaji, 1977a, 23;reprinted in Banaji, 2010, 80).Moreover, in 1980 Banaji was also happy to contribute to Harold Wolpe's edited volume, The Articulation of Modes of Production (Banaji, 1980).
We have seen that, according to Banaji in Theory as History (2010), in any particular society at a particular time the presence of an underlying essential relationship of production, associated with the wage-labour, might be disguised because it is concretely manifested (or appears to the observer) in a variety of different forms, each of which is associated with a particular relationship of exploitation.This is Banaji's principal reason for rejecting the idea of a social formation, which in his opinion is necessarily connected to the erroneous belief that each mode of production is to be associated with just one relationship of exploitation.A similar idea can also be found in Banaji's earlier work, Agrarian Change in Late Antiquity: Gold, Labour and Aristocratic Dominance, which was published in 2001.In that text, however, Banaji expresses this idea in a different way, one which relies on the very concept of a social formation which he later rejects.For example, he asserts there that in Western Europe 'late antiquity throws up a social formation' that was associated with a significant degree of wage-labour, which was 'structured in diverse, flexible ways (as labour tenancy, sharecropping, service contracts, the contracting of specific jobs, etc.).' This, he says, complicates the effort to achieve 'scholarly understanding' of what was happening in European society at that time (Banaji, 2001, 217).On this occasion, therefore, Banaji evidently did not think that use of the concept of a social formation by Marxist historians is an obstacle to achieving an understanding of the diversity and complexity of different relationships of exploitation.It is only since then that he has, for some reason, changed his mind about this issue.
It is also worth drawing attention to Banaji's attitude towards the work of Yevgeny Preobrazhensky.In an article entitled 'Backward Capitalism, Primitive Accumulation and Modes of Production,' which he published in 1973, Banaji refers favourably to Preobrazhensky's view that 'any process of primitive accumulation implies an articulation of modes of production' (Banaji, 1973b, 396).Banaji's reference to Preobrazhensky's implicit use of the concept of a social formation (via the notion of the articulation of modes of production) seems to me to be particularly significant, because he dedicates the essays which are reprinted in Theory as History to the memory of Preobrazhensky and his work.
Although Banaji rejects the concept of a social formation as something which is alien to the outlook of historical materialism as he understands it, nevertheless it is clear that he does wish to retain the notion of a mode of production.In this respect, there is a certain similarity between Banaji's ideas and those of Ellen Meiksins Wood, who argues in a similar way (Meiksins Wood, 2000, 64).In Banaji's view, as in that of Meiksins Wood, the concept of a social formation and that of a mode of production can and should be detached from one another.Like Meiksins Wood, Banaji argues that the concept of a mode of production can be found in Marx's writings, whereas that of a social formation cannot.It is, rather, an alien intrusion or an unnecessary accretion which has (undesirably) been added to Marxism by later Marxists, not least Louis Althusser and his structural Marxist disciples.Against this view, it might be suggested that for Marx the concept of a mode of production and that of a social formation are necessarily related to one another, as a part is related to the totality or whole of which it is a part.As such, they stand or fall together.Once the role that both concepts have to play in Marx's writings has been properly understood, it becomes clear that it is not possible to reject one of them without also rejecting the other.However, I do not see how anybody who wishes to be associated with Marxism could possibly reject the notion of a mode of production.Nor am I suggesting that Banaji would wish to do so, no matter how broadly he wishes to conceive of it.
The essence of my argument has been to suggest that the concept of a social formation is not in fact to be associated with orthodox or vulgar Marxism, as Banaji understands it.Rather, it should be associated with classical Marxism, or with Marxism properly understood, that is to say, with the Marxism of Marx himself and of Lenin, of which Banaji evidently approves.The point of this article has been to subject Banaji's reasons for rejecting the notion of a social formation to an immanent critique.Assuming for the sake of the argument that such a thing as Banaji's 'vulgar Marxism' exists, I can see no decisive reason why it is not possible to sympathise entirely with the criticisms he makes of it, whilst at the same time endorsing rather than rejecting the concept of a social formation.