Lucien Goldmann made important contributions to French Marxist theory and philosophy in the 1960s. Unlike other luminaries like Althusser and Poulantzas, Goldmann took a cautious stance on the strongest claims of scientific certainty for the theses of historical materialism and Marx’s theory of capitalism. Instead, he placed more emphasis on the dialectical core of Marx’s theories — without assuming that Marxism provides a key to understanding the necessary unfolding of history and society. He interprets Marx's thought in terms of the ideas of Hegel's dialectics.
Particularly interesting is his 1970 book, Marxisme et sciences humaines. (Here is a digital edition of the book; link.) Key parts of the epistemological and metaphysical ideas in Goldmann's philosophy of Marxism are also expressed in his essay "Is There a Marxist Sociology?" (link). This piece appeared first in 1957 in French and in an English translation in International Socialism in 1968. In this piece Goldmann considers the debates of the early twentieth century between orthodox Marxists and "ethical" reformist socialists -- between those who believed that socialism was a scientific necessity and those who believed socialism would come about because the masses would come to see that it was the most just social order. Here is Goldmann's summary of his assessment of these debates:
To sum up, what characterises these three fundamental positions (despite their differences, we are classing together Kautsky and Plekhanov) is that they all hold that Marxism implies an objective science distinct and separable from any value judgment, what might be called, to use Poincaré’s terminology, a ‘science in the indicative mood.’ On this point, the different trends of Marxist philosophy merely follow the scientism which characterised academic thought at the end of the nineteenth and the beginning of the twentieth century, by that very fact diverging from the dialectical tradition of the classical German thought of Kant, Hegel and Marx. The differences between these three positions consist in the fact that Vorländer, and with him a large number of thinkers who are partisans of explicit reformism, affirm correctly with Poincaré, that from a science in the indicative mood one can never derive any conclusion in die imperative mood, and thus there could be no ‘scientific socialism,’ since any taking up of a socialist position necessarily has an ethical basis. This position very rapidly became the ideology of a certain explicit reformist trend consisting primarily of some bourgeois democrats brought to socialism by taking seriously the demand for individual freedom for all men.
Goldmann's own position favors the dialectical style of thinking often attributed to Marx's thought, and he rejects the scientistic interpretation associated with orthodox Marxism. In this aspect he affiliates his thinking with that of Georg Lukács. Here is the intellectual swerve that Goldmann most appreciates in Lukács's theories in History and Class Consciousness: Studies in Marxist Dialectics:
Lukács showed that if one accepted the idea of an objective science in the indicative mood, action could no longer be conceived except in terms of ethics or social technology; inversely, if one began with a conception of historical action as individual action, one could conceive it only in ethical or technical terms, and sooner or later, if one developed one’s thought consistently, one would arrive at the idea of an objective science of society. But it is precisely all these complementary concepts – sociology, objective science of social life, technical or ethical action – which seemed to him questionable and above all undialectical.
For him, what characterises historical action, is precisely the fact that it is not carried out by isolated individuals, but by groups who simultaneously know and constitute history. Therefore neither the group nor the individual who is part of it can consider social and historical life from the outside, in an objective fashion. The knowledge of historical and social life is not science but consciousness although it must obviously strive towards the attainment of a rigour and precision comparable to those achieved in an objective fashion by the natural sciences. Any separation of judgments of fact and judgments of value, and, similarly, any separation of theory and practice is impossible in the process of understanding history; the very affirmation of such a separation will have an ideological and distorting effect. Historical knowledge is not a contemplative science; historical action is neither social technique (Machiavelli) nor ethical action (Kant); the two constitute an indivisible whole which is a progressive awareness and the march of humanity towards freedom.
In my own treatment of Marx's ideas in The Scientific Marx (1986) I argued that Marx did not make use of a dialectical method when it came to his social theories.
Here is my summary of my thinking about dialectics in Marx:
It is no doubt true that Marx's mature works contain a certain amount of admittedly Hegelian language and concepts. Marx writes in Capital, "I openly avowed myself the pupil of that mighty thinker [Hegel], and even, here and there in the chapter on the theory of value, coquetted with the mode of expression peculiar to him" (Capital II, pp. 102-3). And in the same passage he speaks with approval of the dialectical method: "The mystification which the dialectic suffers in Hegel's hands by no means prevents him from being the first to present its general forms of motion in a comprehensive and conscious manner. With him it is standing on its head. It must be inverted, in order to discover the rational kernel within the mystical shell." There is thus some fuel for the argument that Capital is not an empirical work but rather a work of materialist philosophy in the Hegelian mode. If these dialectical ideas ran deeply the charge would be compelling. In the following, however, I will argue that Marx is irreconcilably opposed to the use of dialectical logic as a method of inquiry in history or social science. At most the dialectical method represents a highly abstract empirical hypothesis about the nature of social change. I hope therefore to leave the way clear for an interpretation of Marx's scientific method that is in basic agreement with orthodox empirical social science. When Marx goes to work on his detailed treatment of the empirical data of capitalism, he leaves his Hegelian baggage behind. (TSM 113)
Moreover, I note that Marx is sharply critical of Proudhon, in large part because of the appeal that Proudhon makes to the logic of the dialectical method:
Consider finally Marx's critique of Proudhon's political economy in The Poverty of Philosophy. This discussion is particularly important in the present context because Proudhon does attempt to base political economy on a dialectical method of inquiry and explanation, and Marx sharply rejects the possibility of such a science. "What Hegel has done for religion, law, etc., M. Proudhon seeks to do for political economy" (PP, p. 107). Marx describes Proudhon's method in these terms: "If one finds in logical categories the substance of all things, one imagines one has found in the logical formula of movement the absolute method, which not only explains all things, but also implies the movement of things" (PP, p. 107). Thus Proudhon's project is defined as an attempt to assimilate the categories of political economy to an abstract logical system derived from Hegel's Logic, and then to derive the economic laws that can be deduced from this system. Marx's commentary on this approach makes it plain that he thinks it entirely spurious as a technique of scientific inquiry. Here again Marx's critique of dialectics as a speculative, a priori analytic tool is sharp and unforgiving. "The moment we cease to pursue the historical movement of production relations, of which the categories are but the theoretical expression, the moment we want to see in these categories no more than ideas, spontaneous thoughts, independent of real relations, we are forced to attribute the origin of these thoughts to the movement of pure reason. . . . Or, to speak Greek—we have thesis, antithesis and synthesis" (PP, p. 105). "Apply this method to the categories of political economy, and you have the logic and metaphysics of political economy" (PP, p. 108). (TSM 117-118)
So I argue that Marx does not embrace a philosophical method of knowledge discovery, but is instead highly attentive to the constraints of empirical and historical investigation. We need to discover how the social world is rather than how philosophy predicts it should be. And in fact, these commitments allow us to understand Marx's famous comment about "standing Hegel's method on its head":
We are now able to interpret Marx's celebrated remark that with Hegel the dialectic is "standing on its head. It must be inverted, in order to discover the rational kernel within the mystical shell" (Capital II, p. 103). What is the rational kernel, and what is the mystical shell, of the dialectic? And in what sense does Marx "invert" Hegel's method? It is Marx's endorsement of Hegel's view of the historicity of social institutions and their internal dynamics of change that leads him to speak with favor of the dialectical method. "In its rational form . . . [the dialectical method] regards every historically developed form as being in a fluid state, in motion, and therefore grasps its transient aspect as well" (Capital II, p. 103). On my view of Marx's meaning, the rational kernel of Hegel's method is the empirical hypothesis that historical and social processes develop according to an internal dynamic, and that it is possible to provide a rigorous analysis and explanation of historical change based on knowledge of that dynamic. Moreover, Marx plainly accepts Hegel's view that change in history proceeds through substantive contradictions. These theses characterize history as a law-governed process, and one whose changes develop as the result of internal contradictions. Thus the kernel of Hegel's dialectic, on Marx's view, is not methodological at all, but rather a revealing insight into the character of social reality. These theses are empirical hypotheses (albeit formulated at an extremely high-level).
The mystical shell of Hegel's method, by contrast, is methodological—and perniciously so. It is Hegel's belief that pure a priori analysis can allow him to discover the key to this internal dynamic. This assumption is the "logical mysticism" identified by Marx in the Critique of Hegel's "Philosophy of Right," and he emphatically rejects this philosophical method. Finally, the inversion Marx proposes requires that instead of beginning with ideas and attempting to reproduce the material world in thought, we must begin with the material world and attempt to arrive at ideas that adequately describe its real characteristics. "With me the reverse is true: The ideal is nothing but the material world reflected in the mind of man, and translated into forms of thought" (Capital II, p. 102). Thus Marx's inversion of Hegel's method is materialism; more exactly, it is a form of empiricism because it stipulates that knowledge of the material world may be acquired only through detailed, rigorous investigation of concrete empirical and historical circumstances. And what are the methods appropriate to this sort of investigation? They are the methods of empirical science. (119-120)
In 1986, then, I was of the opinion that the matter was open and shut: Marx was not a dialectical thinker. But I now see that there is another way of looking at these issues. If we take the ideas of change, contingency, historical conjunction, and dynamic social processes to be crucial for understanding the social world -- as I do -- then perhaps there is a version of "dialectical thinking" that does not bring along the apriorism that Marx clearly rejects, while at the same time capturing something important about history and social process. The determinism of "historical laws of motion" sounds suspiciously positivistic; whereas the idea of a set of social processes that interact conjuncturally and produce change in often unpredictable ways is more convincing as a description of real history.
So I'll end with an interesting passage from Marxisme et sciences humaines that seems to bear on this question. This is a passage in which Goldmann refers to his own conception of "genetic structuralism", an idea which in turn seems to capture his own view of dialectics:
Du point de vue historique, le structuralisme génétique est apparu, me semble-t-il, pour la premiere fois comm idée fondamentale dans la philosophie, avec Hegel et Marx bien que ni l'un ni l'autre n'aient employé explicitement ce terms. Il n'en reste pas moins que les pensées hégélienne et marxiste sont, pour la premiere fois dans l'histoire de la philosophie, des positions rigoureusement monistes, structuralistes et génétiques. A un niveau immédiat ce phénomene peut etre lié en partie au fait qu'avec Hegel et surtout avec Marx la philosophie moderne se detache progressivement des sciences mathématiques et physiques pour s'orienter en tout premier lieu vers la reflexion sur les fait historiques; il me parait important de constater que, loin d'etre une découverte tardive en sciences historiques et sociales, le structuralisme génétique est au contraire une des premieres positions elaborees par les penseurs qui se sont orientes sérieusement vers un essai de compréhension positive de ces faits. (21-22)
Or, in my poor translation:
Genetic structuralism is one of the first positions created for thinkers who are oriented towards an effort at a positive comprehension of these facts.
This seems to lie at the core of Goldmann's view of how to resolve the debate between orthodox and reform interpretations of Marx: neither positivist science nor ethical pieties, but rather a view of history that envelopes individuals and groups creating their own histories, in circumstances not of their own choosing. And that process seems to be what Goldmann wants to identify as "genetic structuralism" and the outcome of dialectical historical concretization.