The theory of underdevelopment—the only original alternative to the classical growth theories of Smith and Ricardo—is decidedly not evolutionist. As is well known, evolutionism had a major influence in practically all scientific fields. Marx himself harboured a great admiration for Darwin, the formulator of one of the most important scientific paradigms of all time, whose dominance is today near absolute. But neither Marx nor the theorists of underdevelopment were evolutionists. Marx’s theory, which focused on ruptures, saw concrete class interests as the driving force of history—that is, the consciousness, however imperfect, of constituent subjects: ‘Men make their own history’. Evolutionism excludes ‘consciousness’: natural selection operates by chance to eliminate the weakest. For their part, the theorists at the un’s Economic Commission on Latin America (cepal) were influenced by Weber—also at the margins by Marx—whose paradigm is singularity: not selection, but action imbued with meaning. There is no Weberian equivalent of the evolutionary ‘finality’ of the reproduction of the species.

Underdevelopment, then, did not form part of an evolutionary chain stretching from the primitive world through successive stages to full development. Rather it was a historical singularity—the form of capitalist development in ex-colonies, now become a periphery of the world system, that furnished inputs for capital accumulation in the core. This relationship, which persisted even through drastic transformations, was just what prevented the former colonies from ‘evolving’ into the higher stages of capitalist accumulation; that is to say, from catching up with the dynamic centre, however often they received injections of modernization from it. Marxism was equipped with the most formidable arsenal for the critique of classical economics, and possessed a general account of capitalist development in its theory of accumulation. But it failed to specify its concrete historical forms, above all in the periphery. When it did attempt this, it obtained major results—the ‘Prussian road’, ‘passive revolution’—but of the most general sort. Indeed, for a long time a kind of Marxist ‘evolutionism’ held sway, yielding a rickety theorization of the capitalist periphery based on Stalin’s schema of historical stages, running all the way from a primitive communism before the emergence of classes, to a modern communism after their disappearance. In the case of Latin America, where the theory of underdevelopment was considered ‘reformist’ and an ally of us imperialism, stagism led to serious errors of political strategy.

Underdevelopment could be classified as an instance of Gramscian ‘passive revolution’, as Carlos Nelson Coutinho and Luis Jorge Werneck Vianna maintain.footnote1 But unlike the theory of underdevelopment, this notion tells us nothing about the particular ex-colonial conditions of Latin America that give the states of the region their political specificity. Nor does it touch on the descent of labour from the degrading institutions of slavery and the encomienda, which confer on them their social specificity. Florestan Fernandes came close to an interpretation along such Gramscian lines in A Revolução Burguesa no Brasil (1975), but he also owed much to cepal and Celso Furtado. Behind these writers lay the classic analyses of Brazil produced in the 1930s, which dwelt on the peculiarities of Portugal’s colony in South America and a sociability shaped by a combination of the Iberian legacy and a system of exploitation based on slavery.

Underdevelopment was thus not a truncated evolution, but a product of dependency, issuing from the conjunction of Brazil’s place in the international division of capitalist labour with the articulation of domestic economic interests. For this very reason the internal class struggle offered an opening—linked to a shift in the international division of labour—in the shape of the Revolution of 1930 which brought Vargas to power; and the industrialization by import-substitution that ensued from it. In his Formação Econômica do Brasil (1959), Celso Furtado gave us the key to that conjuncture: the crash of 1929 leading to a kind of Brazilian 18th Brumaire, in which industrialization arose as a project for continued domination through other forms of the social division of labour—even at the cost of toppling the coffee-owners from their central position within the local bourgeoisie. The term ‘underdevelopment’ is not neutral: its prefix indicates that the peripheral formations so constituted have a place in the international division of capitalist labour, which is accordingly hierarchical, since otherwise the concept would be meaningless. But the concept is not stagist in either a Darwinian or Stalinist sense.

My ‘Critique of Dualist Reason’ attempted to bring these crossed paths together: as an exercise in critique it belonged to the Marxist tradition, and as a study in specificity to the line of cepal. Although the passions of the time moved me to some invective against the cepalinos, I have long repented of those errors, which were a clumsy way of trying to introduce new considerations into the building of a specifically Brazilian model of underdevelopment—in their fashion, a homage of vice to virtue. The essay was Marxist and cepalino in the sense that it sought to show how the articulation of the economic forms of underdevelopment included political forces, not as an external contingency but a structuring factor. Furtado had touched on this in his interpretation of the crisis of overproduction of coffee in the 1930s, but then abandoned this great insight. The Eighteenth Brumaire should already have taught Marxists that politics is not external to class movements, that classes are forged in struggle; but they too had forgotten this lesson. Here were the two legacies to which I returned in trying to understand why and how leaders like Vargas and their creatures—the Partido Trabalhista Brasileiro (ptb) and the Partido Social-Democrático (psd)—had presided over Brazil’s industrialization, resting a modern manufacturing sector on a backward subsistence agriculture.