Marxism as a Learning Process

“Learn, learn, learn…”

[printable copy]

By Stephen D’Arcy

In everyday justifications offered on behalf of contested claims, simply noting that someone once said something about case X offers little or no help in substantiating a claim about case Y, regardless of how esteemed the speaker may have been. For instance, if I want to know how many people live in Greater Manchester today, it is of very little help to cite the remark by Friedrich Engels, in The Condition of the Working-Class in England in 1844, that Greater Manchester “contains about four hundred thousand inhabitants, rather more than less” (Engels 1892, p. 45). Similarly, albeit less obviously, if I want to determine the nature of a revolutionary process in the 21st century, I would seem not to get very far by quoting Karl Marx’s assertion, made in 1848, that the French Revolution at the end of the 18th century was “the victory of bourgeois property over feudal property, of nationality over provincialism, of competition over the guild, of the partition of estates over primogeniture…, of civil law over privileges of medieval origin.” As interesting as these comments by Engels and Marx may or may not be, they were simply not talking about, and for essential reasons were incapable of making informed judgments about, events or facts occurring over a century after their respective deaths. This much seems utterly uncontroversial. Nevertheless, in marxism, it is common to justify or to criticize claims made in the context of debates today by quoting statements made by Marx, Engels, or other ‘classical’ figures in the marxist tradition, like Lenin or Luxemburg, concerning issues or debates from the 19th or early 20th centuries. These quotations are supposed to serve not simply to enliven one’s writing with memorable aphorisms, but to justify important and contested claims.[1] It is for this reason that this practice is controversial.

Often, critics condemn what they call a “religious” attitude toward Marx (and other figures and texts with — I dare say — “canonical” status). The metaphorical or analogical accusation of “religiosity,” in this context, is a rhetorical means of ascribing to others a certain kind of epistemic irrationality, namely, what the critics deem to be a superstitious stance of deference to ancestral authority. For ease of reference, I refer to this way of objecting to the justificatory deployment of ‘classical’ formulations as ‘anti-citationalism.’ Conversely, I use the term ‘citationalism’ to label the practice of citing classical formulations as if they should be regarded as somehow epistemically authoritative in contemporary debates, in spite of the passage of time and the changing of circumstances.

The anti-citationalist, typically, likes to claim the mantle of rationality and to assume the posture of guardian and guarantor of evidence-based scientificity, determined to hold encroaching unreason at bay. But, notwithstanding its self-assured stance of secular modernity and its air of epistemological sophistication, this resistance toward the citation of classical texts in a justificatory role is misguided, at least in very many cases. In this paper, I want to explain where the anti-citationalist position goes astray.

In order not to miss what is at stake in the citationalism controversy, it is necessary to appreciate upfront that the kind of justification offered by citers of Marx and other classical figures is not empirical, in the sense of offering independent factual evidence to support a claim about what is held to be observably the case. Instead, the citationalist’s appeal is to a judicious precedent, that is, the precedent of a sound judgment offered in an earlier, admittedly different, yet relevantly similar set of circumstances. This appeal has less to do with religious justification (as in, “The Holy Bible says….”) than with common-law juridical justification (as in, “The Court held in a relevantly similar case that the principle of freedom of association was applicable under circumstances of this type….”). A crucial step on the way to vindicating the rationality of citationalism, therefore, is to show that (and in what manner) common law legal reasoning from analogous precedent is — when it is done well, and not poorly — quite capable of qualifying as epistemically rational. 

In this paper, I approach this matter in four steps. First, I draw attention to an assumption that seems to motivate anti-citationalist doubts about justifications of this type, namely, the assumption that the epistemic value of a quotation must be equivalent to what we learn from the quotation itself, from someone having said that. Second, I introduce the idea of justification by recapitulation, which Hegel called “Erinnerung” (recollection) and which is now more commonly known as “rational reconstruction.” According to this idea, one way to justify a claim is to retrace the steps of the learning process by virtue of which we came to be rationally motivated to adopt the claim in the first place. Third, I argue that common-law juridical justification, the legal or jurisprudential variant of the more general phenomenon of precedential reasoning, is rational because and to the extent that it appeals to the discoveries of a cumulative learning process that generates an evolving intellectual inheritance shared by a community of co-inquirers. This inheritance establishes a (defeasible) presumption in favour of ensuring consistency of new judgments with the accumulated stock of previously recognized insights that comprise the tradition. Because cited precedents are presumed to be susceptible to defence by recapitulative justification, the burden of justifying contrary views that break with that inheritance is relatively demanding. Fourth, I draw attention to the fact that the citation of a precedential case in juridical justification can take the form of simply naming the case, as in R. v. Oakes or Roe v. Wade, without explicitly restating all the reasoning or evidence considered in the cited case. This is sufficient, normally, because (by hypothesis) it is a presumptively well-founded established precedent. So too in the citation of canonical (that is, precedential) formulations in marxist social science, the reliance on the short form of citation (that is, merely quoting a single sentence or short passage) implies no deficiency of epistemic rationality, as long as the appeal to learning processes already traversed and susceptible to rational reconstruction can be vindicated on demand. Finally, fifth, having explained the basis for the assumption that the citation of classical formulations is epistemically rational, I illustrate the analysis with two examples drawn from recent debates within marxist social science, in which canonical formulations inherited from the tradition are deployed in a justificatory role, and which are not plausibly seen as irrational in the way that anti-citationalism imagines such justifications to be. In a brief conclusion, I explain why this type of justification is not limited to any particular category of claims, like descriptive, explanatory, or normative claims, but applies to any class of claims that admit of cumulative learning processes to which later inquirers can appeal in the mode of precedential “recollection.”

The Anti-Citationalist’s Motivating Worry

Consider a typical expression of anti-citationalism, to be found in the work of a marxist philosopher of science, Brian Aarons. He writes:  

[M]any marxist intellectuals and revolutionary activists … think that just because Marx or Lenin or Mao wrote something it must therefore be right. For too long this sort of think­ing has led to ‘proof by quotation’ arguments, which are very often futile because the quotations are treated as holy writ, as well as being usually quoted out of the con­text of the text and the times in which they were written.[2]

I note that Aarons describes the citation of classical formulations as “very often futile,” not futile by their very nature. But I am less concerned with replying to this specific formulation than elucidating the type of worry that motivates the anti-citationalist objection as such. Aarons points to three features of the citationalist move: first, the reverential attitude toward the “canonical” source (“holy writ,” true because someone special uttered it); second, the gap in historical context between the original text and the chronologically later citation of it in a different setting; and third, the disembedding of the cited judgment from the context in which it was originally articulated.

It is not actually my intention to push back against these worries on their own terms. I do not, for instance, object to the thought that — other things being equal — it would be epistemically irrational to believe some claim p is true, just because someone deemed to be important or special had long ago uttered the claim that is true. Nor do I deny that the historical context in which some judgment was made could make it epistemically irrational to simply apply the same judgment to questions arising in a very different context. Nor, finally, do I reject the worry that a quoted passage, taken out of its original context, can change its meaning and can go from ‘true’ to ‘false’ by being improperly or carelessly recontextualized. My response to worries of the sort expressed by Aarons is more subtle than a straightforward rejectionof these points. Essentially, my claim is that they grossly understate the epistemological sophistication of the very practice of citing canonical formulations in a justificatory role. Put differently, my objection to Aarons’ criticisms is that they underestimate the resources available for epistemically rational judgment that are already found within the practice of reasoning from cited precedents. To see why, we need to adopt a more curious and perhaps a more charitable approach to analyzing the type of reasoning that the citationalist brings into play.

Dialectical ‘Erinnerungen’: The Epistemic Rationality of Retracing our Steps

The first part of the picture I want to sketch here is the dialectical model of justification via recapitulation, or what Hegel calls “Erinnerung” (recollection).[3] Hegel’s “Erinnerungen” (recollections) are called “rational reconstructions” by the Marxist philosopher of science, Imre Lakatos, and for the most part that’s the term I use here, although the terms roots in Hegel’s conception of inquiry is important to take into account, since the marixist practice of precedential reasoning emerged from the same tradition.[4] This is the model of scientific justification that Marx invoked when he articulated his “après-coup” principle. Marx claims, in the first volume of Capital, specifically in the slightly expanded French edition, that “scientific analysis” of “the forms of human life” must take “a course directly opposite to that of their real [historical] development,” so that the analysis “begins post festum [après coup]…with the results of the process of development ready to hand”[5] in advance. A scientific account of capitalism, for example, begins with what we know capitalism to be, and rationally reconstructs its emergence in retrospect, by retracing the steps of its emergence and elucidating the imperatives to which each of those steps responded. In Marx’s case, however, what is at stake is not justification, but causal explanation. One explains how capitalism followed a certain developmental trajectory by reconstructing the causalimpact of systemic imperatives to which it was, at each stage of its development, responding. Hegel, however, is more concerned with rational reconstruction in the sense of retrospective recapitulation of the rational insights, especially the correction of mistakes, that motivated us to change our minds, or (in many cases) to enrich or complicate our conceptual repertoires or vocabularies, in response to deficiencies of earlier, now-superseded understandings. This Hegelian mode of recapitulation does not just explain our change of mind; it justifies successor views by elucidating how they correct mistakes, or otherwise offer epistemic enhancements, and so embody advances in relation to their predecessors.

Hegel’s books are essentially all rational reconstructions, in this sense, but none more explicitly so than his first and most important book, the Phenomenology of Spirit. There, he begins, après-coup, in Marx’s sense, with what we now know to be true, having already traversed a learning process. Looking back from this epistemically privileged standpoint he justifies this claim to know what we believe to be so by means of a retrospective rational reconstruction, retracing the steps of the learning process, reactivating the insights that propelled us past earlier, now ‘relativized’ steps along the path, which we now know to have been inadequate (what he would call mere relative knowing, as in “it seemed so at the time,” in contrast to “what we now know to be so,” full stop, or so-called ‘absolute’ knowing), because they proved in the course of ongoing learning to be unable to account for insights integrated into their successor stages of a learning process that left them behind. On this model of learning, we move forward by accumulating insights, correcting the deficiencies of predecessor views while integrating what was relatively correct in them, a process Hegel dubs “sublation.”

When I describe Marxism as a learning process, it is to the Hegelian conception of justification by rational reconstruction that I intend to appeal. Marxism is a cumulative learning process in which insights are taken on board. These insights can be invoked later, using the short form of citation typified by the practice here named ‘citationalism’: invoking a presumptively authoritative precedent, as part of a learning process that we — participants in the tradition — have already traversed. 

From the Retracing of our Steps to the Citation of our Precedents

The citation of a canonical judgment inherited from a research tradition in which we claim to participate is not, in and of itself, justification by rational reconstruction, in the dialectical (Hegelian and Marxian) sense reviewed above. In crucial respects, the mere citation of a classical source stops short of recapitulation. In particular, the citational gesture is terse to the point of controversiality. One can hear a debate partner’s invocation of a one- or two-sentence quotation from the Communist Manifesto or Capital and think, “Is that your only argument?” This reaction from the citationalist’s skeptical interlocutor is, if not exactly correct, at least understandable. It responds to the fact that the citationalist’s quotation dislodges a fragment from a larger intellectual and historical context, and then recontextualizes it by inserting it into a different debate, addressing different circumstances, in a different time and place.[6] Extracting two or three sentences from a book like Capital, and recontextualizing them into the back-and-forth of a 21st-century debate, could well appear to be a grossly questionable or even careless procedure, especially if the alleged relevance of the quoted passage extended no further than the sheer worthiness to high esteem attributed to the quoted author. 

However, these doubts are based, in most cases, on a misunderstanding of what the citationalist is up to. The citation of an inherited, canonical judgment is important to us for a very specific reason, not taken into account by the anti-citationalist. We care about the quoted passage, not because the one who uttered it is an object of reverence, but because the passage is the repository or trace (in the sense of footprint) of insights emerging from a learning process that we, who take it as presumptively authoritative, could in principle rationally reconstruct, if called upon to do so, normally by recapitulating a learning process that has already been traversed, probably at the time of the initial formulation (and typically recorded in the larger canonical expression of the cited view). Thus, although the actual citation of the precedential judgment does not contain a justifying recapitulation, the epistemological authoritativeness of the cited precedent is grounded in the availability in principle of a rational reconstruction of its emergence from an insight-motivated learning process. 

We can shed light on the epistemological infrastructure of this practice by considering a parallel in the law. Citationalism in social-scientific research is comparable, epistemologically, to the practice of attributing epistemic authoritativeness to the citation of case law in common-law jurisprudence. The authority of the citation of a previously decided case as precedent, which may take the form of a mere mention of a case’s name (such as, to cite a famous example, Roe v. Wade), or at most an extremely terse summary of a key conclusion, will seem epistemologically mysterious if we lose sight of its relatively complex epistemological structure. There is nothing magical about invoking the name of the decision, nor of the judge or court that generated it. Rather, the case’s claim to our attention in a present-day debate, as an authoritative precedent that is presumptively binding on us in today’s different circumstances, rests on a dual claim that the basis on which the decision was made in the cited case is, first, well-founded by reasoning that is normally recorded in the original decision, or otherwise could be stated or restated by those counting it as precedential, and second, originating as a judgement about an earlier case that is relevantly similar to the one under consideration today. If the best reasoning available to support the supposed precedent were found to be a rationally baseless “hunch,” or some now-discarded theory that was misjudged at the time to be informative, then the presumption in favour of the earlier decision’s bindingness on us would be weak to the point of irrelevance (rationally, if not legally). In just this way, precedents are discarded and overridden over time, as the reasoning on which they rest loses its grip on us. Likewise, if the circumstances under which the decision was made were so different as to cast serious doubt on the applicability or relevance of the precedent to this case, under these circumstances, then once again it would lose its authoritativeness as applied to our situation. If, however, on the contrary, the reasoning is not found wanting or unavailable, and the circumstances are apparently similar in all crucially relevant respects, then the bindingness of the precedent is presumptively dispositive. (The word “presumptively,” of course, implies also the possibility of defeating the presumption; but doing so rationally requires that there be a compelling basis for setting aside the precedent, because there is an epistemological asymmetry between a firmly established precedent, on the one hand, and a possibly eccentric or capricious departure from that established precedent, on the other hand.)

Some Examples

I want to conclude by illustrating this analysis with two simple but vivid examples. These examples illustrate the type of claim that citationalism legitimizes, namely, justification by citation of a judgement recorded in a canonical formulation that encapsulates an insight arrived at via a learning process that the tradition has already traversed. I do not take a position on whether the justifications in these examples are convincing, all things considered. My only claim is that the anti-citationalist objection reviewed above falls flat, when we understand properly what is happening in these appeals to ‘classical’ sources. 

First, consider the case of the debate between John Holloway and Michael Lebowitz in 2005, in the journal Historical Materialism. At issue, centrally, was the anti-statist position of Holloway, according to which the anti-capitalist left should refrain from aiming to “take power,” in the sense of wielding governing authority in the context of the capitalist state.[7] Lebowitz complained that refraining from taking power over the capitalist state was, as he put it, “a profound rejection of Marx.” Holloway, in response, cited a classical passage from Marx, one that Marx not only wrote, in the Eighteenth Brumaire of Louis Bonaparte from 1858, but also quoted — citing his own work — in the Civil War in France from 1871: “the working class cannot simply lay hold of the state machinery and wield it for its own purposes.”[8] On one level, this looks like a debate about who is and who isn’t “profoundly unmarxist,” with Holloway using a quotation from Marx to underline his own claim to a more rigorous form of orthodoxy, confirming that Marx himself endorsed Holloway’s view and rejected that of Lebowitz. However, there is more to the citation than that, because Holloway notes precisely the crucial variable that concerns me in this paper: the citation as a repository of insights gleaned from a learning process already traversed. Exactly in this spirit, Holloway accuses Lebowitz of forgettingsomething already learned. He writes: “[Lebowitz] prefers to forget that [Marx and Engels] revised their understanding of revolution and the state after the experience of the Paris Commune.”[9] In this way, Holloway draws attention to what I earlier called the epistemological infrastructure of the citationalist’s practice: the invocation of a canonical passage from the research tradition to highlight how an interlocutor has forgotten something that has already been learned, and reverted to a retrograde intellectual position, one that has already been surpassed by the community of co-inquirers who share the inheritance of that cumulative learning process. In the sequence of texts invoked by Holloway, a learning process is recorded, and Holloway reminds Lebowitz of a sequence in which a mistake about the capitalist state has already been exposed and corrected.

Holloway’s citationalism, in this instance, purports to be epistemically rational, because it draws attention to an epistemological asymmetry between statism (or capturing the capitalist state) and anti-statism (or ‘smashing’ the capitalist state) in the marxist tradition, such that a special justificatory burden falls on the statist to overcome the justificatory weight, the precedential authoritativeness, of Marx’s reasons for casting doubt on projects of their type on the basis of the experience of the Paris Commune and other accumulated experiences and insights. Superstitious appeals to ancestral authority play no role in this justificatory move.

Consider, now, a second example: the decision of Nisancioglu and Anievas to cite Marx’s Capital on the “chief moments of original accumulation,” as including colonial dispossession, chattel slavery, and other aspects of early-Modern history deemed by so-called “Political Marxists” (like Robert Brenner and Ellen Meiksins Wood) to be exterior to the causal origins of capitalism and excluded from the category of original (ursprünglich) accumulation:

The discovery of gold and silver in America, the expiration, enslavement and entombment in mines of the indigenous population of that continent, the beginnings of the conquest and plunder of India, and the conversion of Africa into a preserve for the commercial hunting of blackskins, are all things which charac- terize the dawn of the era of capitalist production. These idyllic proceedings are the chief moments of primitive accumulation …. The different moments of primitive accumulation can be assigned in particular to Spain, Portugal, Holland, France, and England, in more or less chronological order. These moments are systematically combined together at the end of the seventeenth century in England; the combination embraces the colonies, the national debt, the modern tax system, and the system of protection.[10]

The anti-citationalist presumably regards this as, at best, irrelevant to the justification of the claims made by Nisancioglu and Anievas, and at worst an irrational invocation of Marx as a kind of High Priest or infallible Pope of socialist theory, whose word is law. Indeed, any such gesture — any appeal to Marx as a fount of wisdom, whose word is somehow conclusive — would be irrational, guilty of a fallacious appeal to authority. However, a proper understanding of the citationalist move of Nisancioglu and Anievas in this case will insist that, on the contrary, it is the careless by-passing by some Political Marxists of a precedential judgment of Marx’s, without a sufficiently sound justification for overriding its presumptive authoritativeness, that risks epistemic irrationality by failing to acknowledge and adjust itself to the epistemological asymmetry between a canonical judgment, constituting an established precedent within a still-vital research tradition, and the solitary assessments of a handful of researchers, who seem to some to be reverting to a set of beliefs (Eurocentrism, notably) that were already found to be wanting in the course of a learning process traversed in an earlier stage of scientific development, in this case by Marx himself. Marx’s judgement about the “chief moments of original accumulation” is the repository of an insight-motivated learning process that has been traversed, that could be recapitulated or retraced today, and that addresses a relevantly similar set of facts in comparison to those addressed by the Political Marxist account of the transition. (Here, it is not my intent to claim that the Political Marxists could not answer this objection; I am only trying to clarify the nature of the objection, insofar as it takes a citationalist form.)


I want to conclude with a comment on the generality of the point made here about the epistemic rationality of citationalism, or reasoning from the presumptive authoritativeness of canonical judgments that encapsulate insights accumulated during a cumulative learning process in a research tradition. Does this apply only to one kind, or a few kinds of judgements? For example, does it only apply to strategic judgements, or to explanatory ones, or conceptual points? No. What is justifiable in this way is not restricted by some specific epistemic quality, like being descriptive, explanatory, normative, conceptual, theoretical or interpretive; precedential reasoning can lend support to judgments of any kind. All that is required is that questions under discussion in contemporary debates be relevantly similar to questions that we have already encountered and intellectually processed in our tradition. The present population of Manchester is not relevantly similar to the population of Manchester in the 1840s, and it is for this reason that citing a classical description is uninformative. Identifying a certain strategic orientation deployed a century ago as ‘opportunistic’ might have a much stronger claim to being relevantly similar to a case today, in which the same (or a very similar) orientation is in play, although here too there may have been changes to the circumstances (i.e., relevant dissimilarities) that make new learning necessary. This debate can be conducted in a way that is wholly endogenous to the citationalist procedures set out above, far from necessitating a break with these procedures. 

My aim in this paper is fairly modest. I obviously do not claim that there has never been or could never be an instance of irrational or fallacious appeals to quotations from canonical sources in the marxist tradition. Instead, I claim that the practice of using quotations from canonical sources is not, as such, irrational. If we understand the epistemological infrastructure of the practice — the rational underpinnings of it — we can grasp how these citations appeal to the presumptive authoritativeness of formulations that condense or concisely convey the core of insights that emerged from learning processes that the intellectual tradition of marxism has already traversed. The rational underpinnings of the practice include, first, the neo-Hegelian idea that we can justify a view we now hold by rationally reconstructing the insight-motivated learning process from which it emerged, and second, the convention that these rational reconstructions do not have to be elaborated in each insistence, but can be invoked in a short-form way, such as by citing the name of a decision or a concise statement of it (like a short quotation), as is common in the epistemologically similar practice of precedential reasoning in common law juridical justification. No claim is true just because of who made it; but a claim might be presumptively authoritative in cases where it was made at the end of a learning process, which has not been overtaken by a change in the circumstances or further learning. 


[1] E.g., one might cite a description by Marx of the traits commonly exhibited by investors, in support of the view that such traits are to be expected in the behaviour of an investor today. Or one might cite an analysis by Luxemburg of the dynamics of social-movement escalation, in support of adopting a strategy that anticipates the possibility of an escalation of that kind occuring. Or one might support an explanation by Lenin of the prevalence of pragmatically motivated (“opportunist”) resistance to militancy in labour movements in order to argue for the importance of cultivating an independently organized ‘militant minority’ in working-class movements. It is important to notice that, in such cases, one does not decide the facts of the new case by appeal to the precedential case. Rather, at most, one appeals to the precedent in order to defend a description of a new constellation of facts in terms that were used (in the earlier case) to describe a relevantly similar set of facts. Precedential reasoning is not a method of fact-finding, but a procedure for ascribing presumptive weightiness to well-established ways of describing, analyzing or explaining facts of a certain kind. 

[2] Brian Aarons, “Science or Pseudo-Science: Althusser & Marxism,” Australian Left Review, 1(39), 1973, p. 7.

[3] See especially §808 of G.W.F. Hegel, Phenomenology of Spirit (Oxford: OUP, 1977), trans. A.V. Miller.

[4] Lakatos borrows this expression from Rudolf Carnap, but Lakatos gives it a much more Hegelian interpretation than the neo-Kantian-influenced Carnap had done.

[5] Karl Marx, Capital: A Critique of Political Economy [1867]Volume I, Chapter 1, section 4 (London: Penguin, 1990), p. 168.

[6] In Chapter 4 of The Accumulation of Capital: An Anti-Critique, Luxemburg  calls this “picking a raisin out of the cake,” i.e., extracting a fragment of text that is atypical of or contrary to the context from which it is taken. In English, we also have the expression, “cherry-picking,” often used in a similar way, albeit usually about the selective use of data rather than text fragments. 

[7] Holloway distinguishes, precisely in his debate with Lebowitz, between “taking power” in this sense and the breaking of the capitalist state by forms of working-class self-governance of (what he considers to be) non-statist types, such as workers’ councils or ‘soviets.’ This complication of Holloway’s position seems not quite to be understood by Lebowitz, who does not draw the distinction in this way. See Mike Lebowitz, “Holloway’s Scream: Full of Sound and Fury,” Historical Materialism, Vol. 13 (2005), Issue 4, pp. 217-231.

[8] Marx, quoted by John Holloway, “No,” Historical Materialism, Vol. 13 (2005), Issue 4, p. 279.

[9] Holloway, op. cit., p. 278.

[10] Marx, quoted in Kerem Nisancioglu and Alexander Anievas, How the West Came to Rule: The Geopolitical Origins of Capitalism (London: Pluot, 2015), pp. 25.