Adventures of the Symbolic: Post-Marxism and Radical Democracy

STRUCTURALISM WAS WELL ON ITS WAY toward fully displacing Sartrean existentialism as the regnant intellectual style in France by the early 1960s. François Furet, as we saw in the previous chapter, explained this success as a “dislocation of Marxist dogmatism.”¹ The Soviet invasion of Hungary, Khrushchev’s revelations about Stalinism, the PCF’s waffling on the Algerian war, the apparent waning of working-class activism, particularly the absence of resistance to Charles de Gaulle’s power grab in 1958—these and other disappointments weakened the old attraction of Marxism-Leninism. Even as some disillusioned leftists turned to Lévi-Strauss and others turned to Lacanian psychoanalysis, still others embraced Louis Althusser’s attempt to reinvent Marxism as a structuralist science. Remembering those years from the perspective of a committed political radical, Cornelius Castoriadis wrote, “Those who lived through those times can testify that being a militant at the beginning of the sixties in contact with certain student and university circles in Paris entailed taking a stand against structuralism in general and Althusser in particular.”²

Castoriadis’s magnum opus, The Imaginary Institution of Society (1975), opens with a sustained critique of Marxism, announcing in unequivocal terms his abandonment of Marx. In fact, however, that critique was essentially a summation of criticisms he had already formulated by the late 1950s. When it came to the articulation of his “positive” position in The Imaginary Institution of Society, the arguments and hypotheses that would earn him considerable fame and dominate the rest of his life’s work, he did not develop these in direct confrontation with Marxism. Not Marxism but structuralism emerges in the pages of The Imaginary Institution of Society as the crucial foil for Castoriadis’s own developing theory of the imaginary. Always a combative, polemical thinker, Castoriadis wasted no time in moving from opposition to Marxism to opposition to structuralism, which by the early 1960s had overtaken Marxism as the cutting edge of French intellectual life.

One of the main culprits in perpetuating this rather undifferentiated view is The Imaginary Institution of Society itself. This large book immediately became the touchstone for anyone engaging Castoriadis’s wide-ranging and challenging thought. Yet, behind the semblance of unity, the book actually comprises two documents, the first composed ten years earlier than the second. The first half, “Marxism and Revolutionary Theory,” was written in the early 1960s and originally published in serial form from mid-1964 to mid-1965 in Castoriadis’s journal Socialisme ou Barbarie.⁵ A comparison between the content of this early text and the version printed in The Imaginary Institution of Society essentially confirms Castoriadis’s claim that nothing was changed except for correction of typos and the insertion of some additional footnotes, duly marked so as to distinguish them from the original notes. The only real change—not acknowledged by Castoriadis—is the addition of chapter titles and subtitled sections to “Marxism and Revolutionary Theory,” but that change is not trivial. It lends the text considerable programmatic clarity and strengthens its links to the later sections composed in the early 1970s. The last installment of the serialized text in Socialisme ou Barbarie in the summer of 1965 ends with the promise that the conclusion will follow in the next number; as it turned out, there was no number forthcoming, because the journal and the group reached an impasse that made it impossible to carry on any semblance of collective work. The unfulfilled promise of a conclusion drops out of the 1975 version of “Marxism and Revolutionary Theory,” and instead Castoriadis moves straight to the second part of The Imaginary Institution of Society. The exigencies of the circumstances of Castoriadis’s composition of “Marxism and Revolutionary Theory” are thereby effaced, the fact that it was written on the fly within a militant milieu being torn apart by internal strife at the center of which was Castoriadis’s break with Marxism. Furthermore, the addition of the titles and subtitles and the erasure of the nonconclusion makes the relationship between the earlier and later halves of The Imaginary Institution of Society look more like the logical unfolding of an argument than two attempts to address the same cluster of issues separated by ten years of further theoretical refinement.

Whereas none of Castoriadis’s many commentators have really unpacked this observation, I will take it as my guide in exploring the development of Castoriadis’s thought during the crucial period from 1960 to 1975. In that process of refinement and alteration, Castoriadis’s long critical engagement with structuralism was an important vehicle. And at the center of that engagement was the question of the relationship between the imaginary and the symbolic, which he perceived as a nodal point in his efforts to rethink the radical project.

The journey that found a temporary arrival point in “Marxism and Revolutionary Theory” began, in some respects, at least as early as Castoriadis’s arrival in Paris in 1945. Born in 1922 in Istanbul of Greek parents, and raised in Greece, Castoriadis had joined the Greek Communist Youth in the late 1930s. During the German occupation of Greece, he was a leftwing student in Athens, where discontent with the Communists led him to participate in Trotskyist cells. The dangers of clandestine political activity under Nazi occupation were only slightly lessened when the Germans were driven out in October 1944, because in the ensuing chaos Communists embarked on a murderous campaign against the Trotskyists.⁷ When Castoriadis was offered a bursary from the École française d’Athènes, he knew it was time to leave Greece. Setting sail on the Mataroa, a New Zealand military transport ship carrying a number of other young Greek intellectuals and artists, he reached Paris in late 1945. His intention was to complete a philosophy thesis with René Poirier demonstrating the intrinsic limits of rationalistic philosophical systems, a theme that looks portentous from the vantage point of his later work.⁸ Instead, he became absorbed in the turbulent politics of the French left in the years immediately after the liberation. Already in 1946, Castoriadis and Claude Lefort, the brilliant young student of Maurice Merleau-Ponty, became the leading figures of a critical wing of French Trotskyism known as the Chaulieu-Montal tendency after their cover names, Pierre Chaulieu (Castoriadis) and Claude Montal (Lefort). By late 1948, the Chaulieu-Montal tendency announced its inability to continue with Trotskyism and declared the creation of the group and journal Socialisme ou Barbarie, borrowing its name from a phrase in Rosa Luxemburg’s Junius Pamphlet. In the manifesto that opened the first issue of the journal and in the years to follow, Castoriadis and his comrades developed an ultra-left-libertarian position that was unique in the French political context. They rejected Trotsky’s belief that Stalinism was a temporary aberration in the transition to classless society and attacked the Soviet Union for betraying Marx by developing a highly organized and particularly oppressive form of the “bureaucratic capitalism” that, in a more fragmented version, had come to dominate the West. Rejecting any firm distinctions between these two forms of “capitalism,” they defined the core conflict in capitalism no longer as the struggle between owners of wealth and laborers without property, but as the conflict between “directors” and “executants” in the production process.⁹ Years before the Soviet invasion of Hungary in 1956 produced wider disillusionment with communism, both Castoriadis and Lefort launched withering attacks against fellow-traveling French intellectuals like Jean-Paul Sartre. Tremendously skeptical of the Marxist-Leninist idea of the vanguard role of the revolutionary party, Socialisme ou Barbarie attempted to subordinate their own theoretical work to the inventive power of workers’ agency. Already by the early 1950s, in the face of the authoritarianism of the French Communist Party and the alliance of the Fourth Republic state with the large-scale corporations that were steering France’s postwar economic modernization, Castoriadis redefined socialism as the struggle of people to gain control over their activities, what he called “autogestion” or “self-management.”

Even as late as 1960, Castoriadis aligned his thinking with Marx, declaring that “every revolutionary worthy of the name will always remain a Marxist.”¹⁰ Yet the declaration concealed a process of questioning that had already taken him beyond anything resembling an orthodox Marxist framework. One sees this emerging clearly in the essays ranging from “On the Content of Socialism, I” (1955) to “Modern Capitalism and Revolution” (1960–61). Indeed, “On the Content of Socialism, II” (1957) notes that Socialisme ou Barbarie had already challenged core socialist ideas, including many held by Lenin and even by Marx; yet, he pointedly remarked, “we ourselves have failed to develop the content of our own ideas to the full,” inhibited by “factors that have dominated the evolution of Marxism itself for a century, namely, the enormous dead weight of the ideology of exploiting society, the paralyzing legacy of traditional concepts, and the difficulty of freeing oneself from inherited modes of thought.”¹¹ A number of major themes emerge in these essays, which anticipate the concerns of Castoriadis’s “Marxism and Revolutionary Theory”: first, the vision of history that takes shape here is no longer one of laws and deterministic processes, but rather of human self-creation; second, the question of “institution,” a key element of Castoriadis’s concept of historical creativity that denotes the process whereby a society draws on a well of significations to institute itself as a specific mode and type of human coexistence; third, a new concept of theory that abandons the pursuit of a closed order of truth in favor of “a living theoretical process, from whose womb emerge moments of truth destined to be outstripped.”¹²

Castoriadis’s rapid move from heterodox Marxism to an outright declaration that “Marxism quite simply no longer exists historically as a living theory” produced intolerable strains within Socialisme ou Barbarie.13 Perhaps in an effort to defuse dissent, he insisted that the “Tendency” he now spearheaded “is the organic outcome of the line of development of the review. Indeed, it merely regroups and systematizes ideas already formulated in the review a long time ago.”¹⁴ At a certain level this was true, given the collective nature of the group’s deliberations;¹⁵ but it would be more accurate to say that it mainly systematized ideas that Castoriadis had already formulated; there had in fact always been debate on many points, even if there had been consensus on some basic principles. Castoriadis’s decisive break with Marxism shattered that consensus, for it entailed a shift from quantity to quality that many members could not countenance. Jean-François Lyotard, who had been a member since 1954, led a scission in the name of upholding the commitment to Marxism. In a parting polemic aimed at Castoriadis, Lyotard denounced the new tendency as a form of “existentialism” because of its apparent subjectivism and voluntarism, and he predicted that Castoriadis would soon abandon revolutionary political activity altogether and fall into philosophy.¹⁶ The charge of existentialism was entirely wrong, though Lyotard was not the last to make it; ironically enough, the fall into philosophy characterized Lyotard’s own subsequent development, but it was certainly true of Castoriadis as well.

Indeed, even if articles of the later 1950s anticipated some of the themes of “Marxism and Revolutionary Theory,” what could not have been foreseen was the philosophical tenor of the 1964–65 text. Philippe Gottraux rightly claims that “Marxism and Revolutionary Theory” was not only a rupture with Marxism but also a definitive rupture with the earlier form of Castoriadis’s own writing. Undoubtedly, he remained oriented toward political questions, but, for the first time, he stepped out of the narrow milieu of the ultra-left and engaged in intellectual debate within the larger “intellectual field,” to use Gottraux’s Bourdieu-inspired terms.17 Suddenly, Castoriadis’s discourse burst the confines of an immanent critique based on the experience of working-class movements and capitalist societies and mushroomed into a philosophical reflection in which no question seemed off-limits. Prior to “Marxism and Revolutionary Theory,” Castoriadis later observed, his writing contains “no mention of any philosopher whatsoever.”¹⁸ Now his text is full of references to Hegel, Husserl, Kant, Fichte, and Merleau-Ponty, social theorists like Ruth Benedict, Margaret Mead, and Lévi-Strauss, and the psychoanalysts Freud and Lacan. It turned out that Castoriadis had never stopped reading philosophy and social theory. Just to consider the 1940s, his copious reading list from his student days in Athens include Ernst Cassirer’s Philosophie der symbolischen Formen and works by Max Scheler;¹⁹ he came to Paris intending to do a dissertation in philosophy; in the mid-1940s, he translated much of Hegel’s Wissenschaft der Logik and considered translating Hegel’s Enzyklopädie²⁰; he and Lefort exchanged erudite remarks about Husserl;²¹ and he never ceased to cultivate his deep interest in ancient Greek philosophy. By his own admission, the day job he held for years as an economist for the Organisation for Economic Co-operation and Development afforded him plenty of time to read and write for Socialisme ou Barbarie.

“Marxism and Revolutionary Theory” opens with a recapitulation and extension of the critique of Marx’s economic theory and productivist orientation that Castoriadis had already developed. The opening sections are most striking for the significant deepening of Castoriadis’s argument linking the deterioration of left-wing politics in the twentieth century to the emptying out of the meaning of praxis already implicit in the deterministic logic underpinning Marx’s thought. This in turn led him to a much more comprehensive engagement with the question of historical determinism. Against Marx’s rationalist philosophy of history, which Castoriadis described as an “inverted Hegelianism,” he argued that no causal model could exhaustively account for history. “The non-causal,” he wrote,

appears as behaviour that is not merely “unpredictable” but creative (on the level of individuals, groups, classes or entire societies). It appears not as a simple deviation in relation to an existing type but as the positing of a new type of behaviour, as the institution of a new social rule, as the invention of a new object or a new form—in short, as an emergence or a production which cannot be deduced on the basis of a previous situation, as a conclusion that goes beyond the premises or as the positing of new premises. … History cannot be thought in accordance with the determinist schema (nor, moreover, in accordance with a simple “dialectical” schema) because it is the domain of creation.²²

Castoriadis linked historical creation to the process of “signification,” whereby “men give their individual and collective life a signification that is not preassigned, a signification that they have to make while they are at grips with real conditions, which neither exclude nor guarantee the accomplishment of their projects.”23

The most novel dimension of “Marxism and Revolutionary Theory” is Castoriadis’s effort to tie these ideas of creativity and institution to a new theory of the imaginary. There is a certain logic to Castoriadis’s arrival at this concept, for having described history as a domain of creation, he next searched for its source. In his earlier writings, this source was simply practice, which he believed is never fully dictated by existing rules and procedures and produces in the aggregate an indeterminate openness to novelty. In shifting so emphatically to a philosophical register in “Marxism and Revolutionary Theory,” Castoriadis seemed to respond to the need for an even more fundamental, perhaps even foundational source of creativity. A central feature of his theory of the imaginary is that it functions at two intersecting, yet distinct levels, that of society and that of the individual subject. In a sense reminiscent of Durkheim, Castoriadis designated the social-historical a sui generis domain that is “neither the unending addition of intersubjective networks” nor “their simple ‘product.’”24 The social-historical is, rather, “a dimension of the collective and the anonymous,” which “is what is everyone and what is no one, what is never absent and almost never present as such, a non-being that is more real than any being, that in which we are immersed yet which we can never apprehend in ‘person.’”²⁵ Castoriadis later compressed these various attributes into a succinct definition of the social imaginary: “Imaginary: an unmotivated creation, that exists only in and through the positing of images. Social: inconceivable as the work or the product of an individual or a host of individuals (the individual is a social institution), underivable from the psyche in itself as such.”²⁶ If social significations are not traceable to the individual, neither is the individual imagination exhaustively traceable to the social-historical. This domain of individual imagination he called “radical imagination,” signifying the power to make images arise in consciousness. We shall see shortly that Castoriadis’s first formulation of the concept of radical imagination owed a great deal to phenomenology, whereas his subsequent formulations drew more heavily on psychoanalysis to describe radical imagination as a continual flux of representations, affects, and intentions that are virtually coextensive with the individual psyche per se. As he wrote in the later half of The Imaginary Institution of Society, the psychical and the social-historical are two domains of the radical imaginary, not reducible to each other but articulated together, not a mediated unity, but a “non-empty intersection between the private world and the public world.”²⁷ “A full recognition of the radical imagination is possible,” he claimed in 1978, “only if it goes hand in hand with the discovery of the other dimension of the radical imaginary, the social-historical imaginary, instituting society as source of ontological creation deploying itself as history.”²⁸

Castoriadis’s theory of the imaginary is original, but it must be said that the term was in the air. There were, first of all, the surrealists, with their idea of the power of poetry to remake the world; André Breton fascinated the young Castoriadis, and two young surrealist poets, Benjamin Peret and Jean-Jacques Lebel, had considerable involvement with Socialisme ou Barbarie.²⁹ Jean-Paul Sartre had done a great deal to put the imaginary on the French philosophical agenda with his 1940 book L’imaginaire, which in turn was strongly indebted to Edmund Husserl. Gaston Bachelard also drew heavily on phenomenology in his works on imagination as well as on both Freudian and Jungian psychoanalysis. Bachelard’s student Gilbert Durand further developed the implications of Bachelard’s theory of archetypes of imagination in his 1963 book Les structures anthropologiques de l’imaginaire: introduction à l’archétypologie générale. Castoriadis studied Bachelard when he first came to Paris but rapidly grew dissatisfied with him.³⁰ While the grounds of this dissatisfaction in the 1940s are not entirely clear, it is not difficult to see differences from the perspective of Castoriadis’s developed theory of the imaginary. Bachelard, and even more Durand aimed to classify the symbolic manifestations of the human imagination as these clustered around certain fundamental structures deriving from both human physiology and the human encounter with the world. While neither Bachelard nor Durand followed Carl Jung into the theory of a collective unconscious, their explorations of the symbolic forms of the imagination bore a timeless, archetypal quality.³¹ By contrast, Castoriadis’s concern was with the origin of images and symbols in a dynamic and endlessly creative imagination.³²

His approach bears a somewhat closer relationship to Sartre’s phenomenological path, but here Castoriadis is not the most reliable guide on the most crucial difference between himself and Sartre. In The Imaginary Sartre followed Husserl in contrasting perception to imagining. Whereas in perception consciousness observes an object that surpasses it, consciousness, in imagination, gives itself an object all at once. The object of perception exceeds consciousness, whereas “the object of an image is never anything more than the consciousness one has of it.” Perception opens toward an infinitely rich world; imagination suffers an “essential poverty.”33 This removal of the image from the perceptual plenum points to a further attribute of the Sartrean imagination, namely that acts of creation are in fact negations of the existent, acts of positing a “thesis of irreality.”³⁴ This view ultimately opens directly onto basic themes of Sartrean existentialism, for even though the imaginary suffers an “essential poverty” compared to the fullness of perception, the imagination is nonetheless the basis of our existential freedom because it is the source of man’s transcendence of the real; though the imaginary may be the source of enslavement to our own fantasies, it is also the source of the spontaneous freedom of consciousness.

Castoriadis insisted on a firm distinction between this “purely negative” view of the imagination and his own creative view.³⁵ His imaginary is an “originary faculty to pose or give oneself, under the mode of representation, a thing and a relation that are not (things that are not or never have been given in perception).”³⁶ In a 1996 interview, Peter Dews and Peter Osborne rightly pressed him on this distinction, asking, “Isn’t the philosophical structure of that process actually the same, with one side rather than the other being emphasized?”³⁷ Indeed, Castoriadis himself recognized that negation belongs essentially to the act of imagination, and he was as ready as Sartre to regard the capacity to negate as a crucial dimension of freedom.³⁸ Conversely, Sartre saw positing and negating as two sides of the same coin: “So to posit the world as world and to ‘nihilate’ it are one and the same thing.”³⁹ Contrary to Castoriadis’s depiction, imagining is in fact a good example of Sartre’s dialectical, as opposed to purely phenomenological, style of analysis, wherein opposites are brought into play rather than bracketed as they would be in Husserl.⁴⁰

The more significant difference to Sartre comes in Castoriadis’s immediate response to Dews and Osborne’s question, when he says, “there is no given without imagination.”41 That is, Castoriadis maintained that the imagination institutes “reality.” Hence, for example, in “Marxism and Revolutionary Theory,” he rebutted the orthodox Marxian determinism of the economic base by claiming, “The ‘real social relations’ concerned here are always instituted.”⁴² In the latter half of The Imaginary Institution of Society, he took this further when he claimed, “society as a whole and every society in particular constitutes the ‘real’ and its own ‘real.’”⁴³ Sartre, by contrast, upheld a division between the real and the imaginary that rests on his initial division between perception and imagination. Here Sartre revealed the extent of his link to Husserl, for Husserl maintained that there is an “unbridgeable and essential difference” between “perception, on the one hand, and on the other, presentation in the form of an image.”⁴⁴ As Richard Kearney writes, Husserl privileged “the perceptual mode of intentionality … because of [its] direct access to the flesh-and-blood presentness of things.”⁴⁵ Castoriadis clearly took something from both Husserl and Sartre insofar as he embraced the notion of intentional consciousness, depicting the presentation of an image as an act, not an object of consciousness. Nonetheless, the form of realism evident in Sartre and Husserl’s distinction between perception and imagination was one that Castoriadis could not accept.

In formulating his theory of the imagination, Castoriadis reached even further into the past, behind phenomenology back to Kant and Fichte. Kant, after all, had developed a theory of the transcendental imagination to account for the schemata whereby categories are applied to given objects. Yet, as Castoriadis argued in a later essay, Kant’s transcendental imagination always produces the “Stable and the Same. There is nothing more deprived of imagination than the transcendental imagination of Kant.” Indeed, Kant’s epistemological framework would collapse if the transcendental imagination were at all capable of creating.46 Not surprisingly, Castoriadis declared his greater affinity for Fichte, at least the Fichte of the first Wissenschaftslehre (1794/95), where the “productive imagination” (produktive Einbildungskraft) not only supplies schemata, as it did in Kant’s first critique, but also produces the intuited objects themselves.⁴⁷ Fichte’s radicalization of Kant’s theory of the imagination is clearly described in the one pertinent source that Castoriadis cites in “Marxism and Revolutionary Theory,” Richard Kroner’s classic Von Kant bis Hegel.⁴⁸ Though Castoriadis never wrote extensively on Fichte, he clearly had an ongoing engagement with his thought. His unfulfilled plans for a magnum opus titled L’ élément imaginaire called for studies of the Sturm und Drang, Fichte, the young Hegel and Schelling, and the Jena Romantics; his unpublished working notes during the 1960s and 1970s contain many references to Fichte.

The influence of Fichte may, if anything, have grown in the years after “Marxism and Revolutionary Theory,” as Castoriadis developed in greater detail his understanding of the psyche as “above all, a ceaseless surging of representations and the unique mode in which this representational flux exists.”⁴⁹ This continual, unmasterable, and spontaneous flux of representations, desires, and affects is creative, in the truest sense of the word, because it begins with the emergence of representation itself. It is impossible not to hear in this “capacity to produce an ‘initial’ representation” an echo of Fichte’s basic move when he addressed the impasses of K. L. Reinhold’s attempt to create a philosophical science of “representations” based on the foundational “fact” (Tatsache) of self-consciousness. As Jerrold Seigel describes the problem, Fichte recognized there is a “basic incoherence in the notion that there is a self that can first come to know itself through conscious reflection: in order to recognize itself in the mirror it holds up, it must first know who it is.”⁵⁰ In order to explain how the self can recognize a certain object as itself, there must already be a prereflective form of self-knowledge. This led Fichte to argue that the first principle is not a “fact” (Tatsache), but an “act” (Tathandlung) whereby the “I” posits itself. Fichte thereby shifted the conception of self-knowledge from one of relation to one of production. The imagination plays a crucial role here, for imagination must produce an “image” (Bild) of the self’s own productive and self-determining activity if consciousness is to recognize itself as the source of this activity.⁵¹ As we will see, a similar theoretical structure underlies Castoriadis’s rejection of Jacques Lacan’s theory of specularity.

Castoriadis repeatedly emphasized a crucial divide between himself and all earlier theorists of the imagination. Namely, where they remained at the level of the subjective imagination, his theory moved between the two poles of the individual subject and the social-historical world.52 These poles, we have seen, are irreducible, and in this sense Castoriadis was a dualist. However, insofar as the subjective and the social meet in a “nonempty intersection,” this junction is itself an “institution.” Institution was the key concept allowing Castoriadis to link the two dimensions of his theory of the imaginary. Moreover, it played a vital role in his conception of autonomy. A constitutive tension runs through his view of the social-historical between the “instituted imaginary” and the “instituting imaginary”: “on the one hand, given structures, “materialized” institutions and works, whether these be material or not; and, on the other hand, that which structures, institutes, materializes. In short, it is the union and the tension of instituted and of instituting society, of history made and of history in the making.”⁵³ Every human society is self-instituted, but generally institution happens behind people’s backs, a product of the anonymous activity of the collective. Societies have typically repressed their instituting activity, assigning their instituted forms to extrasocial sources.⁵⁴ The paradox of self-institution is, therefore, that generally it has instituted society as heteronomy. Autonomous society demands a deepening recognition of the self-instituting activity of society, an open-ended interrogation of the instituted forms of society, and a willingness to change those forms when they cease to serve our needs.

In Castoriadis’s capacious concept of “institution,” the double meaning of the word is constantly in play, referring to both social forms and the act of creating them, but, beyond obvious social institutions as in church, hospital, university, or government, institution implies a certain kind of human being, a certain kind of world, and a certain kind of interaction between world and man. Institution is, to use a phrase of Maurice Merleau-Ponty, a mise en forme du monde, an articulation of the world. It is very difficult to resist turning to Merleau-Ponty in trying to get some hold on Castoriadis’s theory. Curiously, Castoriadis later denied that Merleau-Ponty had anything to do with his attempt to understand the phenomenon of institution.55 There can be no doubt about Castoriadis’s general familiarity with Merleau-Ponty, however. The cofounder of Socialisme ou Barbarie was, after all, Claude Lefort, one of Merleau-Ponty’s most brilliant students. Castoriadis himself wrote two lengthy and detailed essays on Merleau-Ponty in the course of his lifetime.⁵⁶ Merleau-Ponty’s influence seems evident in many aspects of “Marxism and Revolutionary Theory,” for example, in Castoriadis’s discussion of embodiment as a counter to the “point-like ego” of the Cartesian tradition, and nowhere so clearly as in the theory of institution itself. Merleau-Ponty devoted his Collège de France lectures in 1954–55 to the problem of institution.⁵⁷ Institution served Merleau-Ponty as a way to shift his critique of subject-centered philosophy from the account of individual being-in-the-world found in The Phenomenology of Perception to an account of the social-historical mode of human being. He replaced the notion of the “constituting subject” with an “instituted and instituting” subject, a subject that is no longer the source of meaning and coherence, but rather finds itself in a framework or space of meaning that transcends it, even if the subject may help to produce the institution and certainly reactivates it through actions and speech.⁵⁸ To illustrate institution, Merleau-Ponty quoted Goethe, who said that genius is posthumous productivity: just as genius creates a new frame that will provoke and guide future work, so too the institution “sets on course an activity, a succession, initiation into a present which is productive after it.”⁵⁹ Reinforcing the significance of this for the question of the human subject, Merleau-Ponty emphasized that there is no “break” (coupure) between private and public institution, and he refused to assign a causal direction when speaking of the private and the public.⁶⁰

The overlaps with Castoriadis are patent, and it is hard to believe Castoriadis’s claim that Merleau-Ponty had nothing to do with his turn to the topic of institution. Moreover, it is strange to see Castoriadis insist in his interview with Dews and Osborne that, “there is no idea of creation or creativity in Merleau-Ponty, as far as I can see.”61 After all, in the lectures of 1954–55 Merleau-Ponty spoke of revolution as a “reinstitution” and even “reversal” of the previous institution.⁶² Still more fundamentally, even in The Phenomenology of Perception Merleau-Ponty ascribed a creative dimension to perception itself, and Castoriadis himself noted that for Merleau-Ponty perception is “instituted,” even if he also believed that Merleau-Ponty’s process of self-liberation from the primacy of perception had been cut short by his premature death.⁶³ It may be that Castoriadis’s claim was an instance of faulty memory—in the same interview he noted candidly how difficult it is to sort out one’s own intellectual autobiography. It may also be that Merleau-Ponty’s endorsement of Saussurean linguistics and his move in the 1950s toward Lévi-Strauss as an ally in his challenge to the philosophy of the subject may have contributed to Castoriadis’s assessment. Finally, although Merleau-Ponty clearly reserved space for deliberate human action, he was sufficiently Heideggerean that in the equation between human activity and human receptivity, between the act of institution and institution as a new unveiling of being, the latter usually dominated. Castoriadis’s values thus stood in some tension with Merleau-Ponty, and, as we shall see in chapter 4, this would become a deep source of division between him and his erstwhile Socialisme ou Barbarie comrade Claude Lefort, who closely followed Merleau-Ponty’s attempt to embed the conscious subject in a background that it neither creates nor explicitly thematizes.

Merleau-Ponty was in fact never Castoriadis’s critical target; indeed, he always retained a respectful and appreciative tone toward the great French phenomenologist. Instead, his formidable polemical energies moved quickly from Marxism to structuralism; to defend the idea of a radical project based on the pursuit of autonomy and the power of the imaginary in the early 1960s necessarily implied a fight against structuralism. The nodal point of Castoriadis’s struggle against structuralism was the status of the symbolic, which was the heart of the structuralist theory of society. Undoubtedly, matters are complicated by Castoriadis’s acceptance of certain aspects of the symbolic turn most prominently represented by Claude Lévi-Strauss.⁶⁴ Yet shared terrain immediately became divided again. The structuralists were interested in the process whereby cultural systems could be dissolved and reassembled into various codes or constituent elements according to the regular operations of structural laws. Lévi-Strauss’s founding gesture repeated, or perhaps exaggerated, Ferdinand Saussure’s exclusion of the temporal, and his work erased all question of origin by suggesting the fiction that the symbolic order must have appeared all at once.⁶⁵ Castoriadis, by contrast, searched for an understanding of symbolic activity compatible with his vision of social-historical creativity and the struggle for autonomy. Political stakes are never distant from his interrogation. Thus the final page of “Marxism and Revolutionary Theory” reverses Marx’s “thesis eleven” to insist that we interpret the world in order to change it.⁶⁶ Accordingly, his interrogation aimed at the origins of the symbolic and the processes whereby new meaning emerges. To put it in Castoriadis’s own terms, the question of the symbolic is inseparable from the question of institution.

A section titled “The Institution and the Symbolic” begins by emphasizing, “Everything that is presented to us in the social-historical world is inextricably tied to the symbolic.” Although things, acts, and institutions are not reducible to the symbolic, Castoriadis pointedly emphasized that none of these things would be possible outside of a symbolic network.67 Castoriadis’s crucial departure from structuralism lies in his insistence that “something else” is involved in symbolism. “The determinations of the symbolic,” he wrote, “do not exhaust its substance. An essential, and, for our purposes, decisive component remains: the imaginary component of every symbol and of every symbolism.”⁶⁸ Where Lévi-Strauss defines the “symbolic function” as a combination of diacritical elements within a structure, Castoriadis identifies the “symbolic function” with the “imaginative function,” because the basic capacity of symbolism is identical to that of the imagination. That is, both presuppose a capacity to see and to think in a thing something that it is not.

Castoriadis’s emphasis on the imaginary element committed him to a theory of the symbol as an expression of meaning. Castoriadis saw that Lévi-Strauss was, by contrast, increasingly tempted to reduce institutions to symbolic networks in which “meaning is always the result of the combination of elements which are not in themselves meaningful.”⁷⁰

This emphasis on the symbol’s expressive power did not force Castoriadis to regress behind the insights of the “linguistic turn.” He flatly rejected the idea of “expression” in the sense of a preexisting constituted meaning coming into form in language or other symbolic media. There is no “pure thought,” but, just as equally, there is no “pure sign”; thinking and speaking are inseparable, and the “philosophy of a sovereign constituting consciousness” is just as mythic as “a structuralist or semiotic ideology, which takes account only of collections of arbitrary characteristics from which a combinatory would extract some will-o’-the-wisp meaning.”72 As Hans Joas has said, even as Castoriadis aimed to overcome the philosophy of consciousness, he also retained the seriousness of the struggle to express meaning—the process of “articulation”—as crucial to the task of securing “the possibility of novelty and creation against the view of a linguistically closed universe.”⁷³

Castoriadis’s theory opens toward a hermeneutic impulse absent in Lévi-Strauss. Hence, wrote Castoriadis, “Understanding, and even grasping, the symbolism of a society is grasping the significations that it carries. These significations appear only as they are carried by signifying structures; but this does not mean that they can be reduced to these, that they result from them in a univocal manner, or, finally, that they are determined by them.”⁷⁴ This position brings Castoriadis’s theoretical statements back into communication with his political commitment to the transformation of society. For even though people communicate and cooperate in a symbolic milieu, this symbolism is itself created.

“Marxism and Revolutionary Theory” attempts to conceptualize the symbolic constitution of society without lapsing into the polarity that defines much of later twentieth-century thought: the image of a fully present, conscious subject, on the one side, and the image of the subject as an empty intersection traversed by the traffic of signs, on the other. To bring the imaginary and the symbolic into close communication, without collapsing the one into the other, is to acknowledge the power and indispensability of symbolic forms while preserving an element of individual and social life that intertwines with symbolic networks but is not identical with them. To insist on this principle of difference, which is not merely the diacritical difference emerging from structure, is to insist on the open and flexible relationship between our signifying systems and ourselves as meaning-creating subjects, to insist equally on the role of symbols in determining domains of social life and the capacity for freedom and improvisation. At the same time, Castoriadis avoided two possible relapses. First, he was attentive to the process whereby meaning gains clarity and form in the process of articulation. There is no presymbolic meaning, but there are representations, affects, and intentions that are not reducible to the symbolic either in their form or origin. Second, while he thus avoided positing a fully present subject, he also avoided a simple idea of referentiality. Symbols refer, but in contrast to Durkheim or the Anglo-American functionalists, they do not refer to “social reality,” the “natural world” or functional needs. Rather, they refer to imaginary significations, meanings that themselves are not reducible to “reality,” but are, rather, active in the institution of social reality.

This way of thinking about symbols circles back to an important distinction between expressive and semiotic theories of symbolism or, to use Bourdieu’s terms, between symbolic systems as “structuring structures” and symbolic systems as “structured structures.”76 We can take up this distinction in yet another formulation, that of Karl-Siegbert Rehberg from his long introduction to the volume Institutionalität und Symbolisierung. Rehberg distinguishes between Präsenzsymbolik (symbolic of presence) and the Zeichentheoretische Tradition (semiotic tradition). In the first tradition the symbol is not merely a sign; rather, it makes the absent thing present, incarnates it. Typically, this is the idea of the symbol as the disclosure or revelation of a different, higher order; it is linked most closely with the German Romantics and their enchanted vision of the world. By contrast, the semiotic tradition belongs to the disenchanted, rational world. “Today,” writes Rehberg, “the arbitrariness of signs and symbolizing complexes is near to us; distant is the symbol’s magical double existence as presence of something other, as well as its authoritative structure. Compared to the older system—based on Ur-texts and religious conviction—of referral to the invisible, the doctrine of signs proves itself more modern.”⁷⁷

Castoriadis’s approach shows the inadequacy of Rehberg’s division. His distinction between symbols and imaginary significations underscores that symbols stand in close relationship to social meanings that are themselves not symbols but, rather, the stuff of symbolization. This is, admittedly, a difficult position. If one is to avoid a regress that is symbolic all the way down, one must accept that the imaginary presents a dimension of obscurity that may be significant, even constitutive, but is like the dark navel of a dream described by Freud as the limit of dream interpretation. Castoriadis himself recognized the difficulty of trying to grasp imaginary significations as such, when their accessibility is so entwined with symbols, and their own “mode of being” is a “mode of non-being.”⁷⁸ Yet if one is to develop anything like a hermeneutic of social symbols and, even more importantly, a theory based on the emergence of novel meaning, it seems necessary to embrace a theory in which symbols refer to something other than more symbols in a symbolic matrix. However, because Castoriadis wished to break from the givenness of “reality,” the primacy of perception, and the hold of functionalist rationality, the system of referral must open toward things that are not there. Hence, Castoriadis’s theory of the symbolic returns to the dimension of presence dismissed as unmodern by Rehberg and, beyond him, figures like Walter Benjamin, Paul de Man, and Judith Butler.

One already detects this move in “Marxism and Revolutionary Theory,” but it becomes fully explicit in the later half of The Imaginary Institution of Society. There one of the most frequent terms is the verb presentify, along with its noun form presentification. Toward the close of the book, incarnation also enters the discussion. Alluding to Durkheim’s terminology, Castoriadis wrote: “It was found necessary to affirm that social facts are not things. In truth, what is to be said is that social things are not “things”; they are social things and these particular things only inasmuch as they “incarnate” or, better, figure and presentify, social significations.”79 The symbolic network thus has a vital relationship to the invisible, making it present and incarnate. Castoriadis noted pointedly that Marx revealed an awareness of this in his analysis of commodity fetishism, but whereas Marx thought this “hieroglyphic” character belonged strictly to the capitalist mode and would evaporate with the advent of socialism, Castoriadis saw it as a general feature of the social-historical world.⁸⁰ Yet in contrast to Romantic theory, Castoriadis’s hieroglyphic refers not to what Novalis called a “numinous” realm, but to a magma of significations that forms “an indefinite skein of interminable referrals to something other than (than what would appear to be stated directly).”⁸¹ Moreover, “presence” no longer implies a quasi-theological manifestation of the absent. Rather, Castoriadis mobilizes the language of phenomenology for a description of the act whereby consciousness spontaneously makes an image or mental object present. For this phenomenological terminology, one could go back at least as far as Kasimir Twardowski’s On the Content and Object of Presentations (1894) and Husserl’s Philosophy of Arithmetic (1891) with its distinctions between “intuitive” and “symbolic” presentations and its claim that the latter depends on the former.

The representative flux is, makes itself, as self-alteration, the incessant emergence of the other in and through the positing (Vorstellung) of images or figures, an imaging which unfolds, brings into being and constantly actualizes what appears retrospectively, to reflective analysis, as the pre-existing conditions of its possibility. … Obviously, some representations are: for example, perceptive representations, called perceptions, the putting-into-images of … (something about which nothing can be said except in and through another representation). Here, it will be forever impossible absolutely to separate what comes from that which is put into images and what puts into images, the radical imagination, the representative flux. In the same way, representation (Vorstellung) is not re-presentation (Vertretung); it is not there for something else or in place of something else, to re-present it a second time.⁸³

Castoriadis clearly realized that such formulations risked falling into “fictions of a thought without language, a transcendental language or language as condition external to thinking.”84 Here, again, it is important to repeat that even if Castoriadis denied the identity of meaning and forms of representation, he did not believe we have access to meaning outside our modes of presentation (Vorstellung) and re-presentation (Vertretung). This mutual inherence finds its clearest articulation in the following:

“Marxism and Revolutionary Theory” represented a dramatic shift in tone and conceptualization measured against Castoriadis’s earlier Socialisme ou Barbarie writings. Yet, as he reported, striking as his innovations were, they rapidly evolved from “arrival points” to “starting points.” If we compare the 1964–65 sections of The Imaginary Institution of Society to the sections first published in 1975, at least two major developments become apparent.

Incidentally, this effort to avoid subjective idealism points to yet another way in which Castoriadis’s project drew inspiration from Fichte, whose own “realist” moment has attracted increasing attention of late.⁸⁹ This refers to Fichte’s description of the ego’s encounters with a world whose capacity to resist human intentions can never be fully overcome; the recalcitrance of the world is the source of the Anstoß, the “check” and/or “stimulus” that the thing-in-itself presents to the ego’s activity. Castoriadis suggests something similar when he wrote in the latter half of The Imaginary Institution of Society that the world is neither devoured nor created by human significations. The world—the real—lends itself to signification, but it is also “an inexhaustible supply of otherness, and … an irreducible challenge to every established signification.”⁹⁰ The Fichtean element becomes entirely clear in a remarkable text from 1986, “The State of the Subject Today.” There Castoriadis began not with the conscious ego but with the organism in an effort to delineate the dynamics of closure and openness at work in the constitution of the individual entity. Speaking of the X that is the outside of the “for-itself” of the organism, Castoriadis wrote: “all we can say about it is that it creates a shock (Anstoss, to take up Fichte’s term) which sets in motion the formative (imaging/imagining, presenting, and relating) capacities of the living being.”⁹¹ The Fichtean Anstoß must be understood as a companion to Castoriadis’s use of a Freudian term, Anlehnung, to signify the way in which both the psyche and society “lean on” or “borrow” from a reality—whether in the form of nature or the organic basis of the psyche—that transcends both. Were there no such relationship, we would never meet anything but our own representations, an omnipotence of thought reserved for the psychotic.

The second major development after 1965, and the one that is of greatest relevance to the present discussion, is a marked turn toward psychoanalysis. Freudian influences are, of course, obvious in “Marxism and Revolutionary Theory.” Clearly, Castoriadis was reading seriously in Freudian literature by the early 1960s, and he had entered analysis with Dr. Irène Perrier-Roublef in 1960.92 But the theory of the imaginary he elaborated in “Marxism and Revolutionary Theory” owes at least as much, if not more, to phenomenology and German idealism as it does to psychoanalysis. Indeed, although in 1964–65 Castoriadis already insisted on the nonidentity of the social imaginary and the radical imagination of the individual psyche, his ideas about the psyche were in fact relatively undeveloped. There is hardly any discussion of the dynamics of the psyche and how they might support his claims for the centrality of imagination; moreover, Castoriadis offers little on what actually happens in what he later called the “non-empty intersection” between the social-historical and the psychical. In search of a more adequate theory, Castoriadis turned to a much more serious study of psychoanalysis. And that meant, inevitably, a reckoning with Jacques Lacan. For, in 1960s France, Lacan’s concepts of the symbolic and the imaginary provided the dominant framework for posing the question of the relationship between private fantasy and representations, on the one side, and social meanings, on the other. With Castoriadis’s psychoanalytic turn, his campaign against structuralism shifted decisively from Lévi-Strauss to Lacan. Where his critique of Lévi-Strauss aimed to overcome the exclusion of the imaginary from the symbolic, his critique of Lacan attempted to preserve the imaginary as a dynamic productive force against Lacan’s reduction of the imaginary to a structure of capture, fixation, and compensatory fantasy.

Castoriadis entered the world of French psychoanalysis in a particularly fraught phase of its rancorous history. Ever the enfant terrible, Jacques Lacan was expelled from the Société psychanalytique de Paris in 1936, and during the 1950s the International Psychoanalytic Association made various attempts to expel him because of his unorthodox training practices. That conflict came to a head in 1963 when the French Psychoanalytic Society, formed in 1953 by analysts driven out of the Société psychanalytique de Paris, removed Lacan from its list of approved training analysts. In reward, the international association recognized the French Psychoanalytic Society as the only authorized psychoanalytic organization in France. Lacan responded in June 1964 by creating the École Freudienne de Paris. Lacan transferred his weekly seminars from Sainte-Anne Hospital to the École Normale Supérieure, and there the medical people in his audience dwindled even as the ranks of philosophers, anthropologists, linguists, and literary critics swelled. The École Freudienne was itself to be a psychoanalytic society with a difference: no analytic hierarchy, no closed circle of analysts, but rather “a meeting place for the freest possible contact between psychoanalysts and members of other disciplines.”⁹⁵ Lacan’s excommunication and his dramatic entry into the broader intellectual milieus of Paris ensured him what Elisabeth Roudinesco calls a “paradoxical position, at once marginal and essential, to French academic life.”⁹⁶

Castoriadis was intimately related to these events. In 1964 he became a member of the École Freudienne, and his archive contains invitations to attend Lacan’s seminar as well as notes on lectures by Lacan, André Green, Serge Leclaire, and Piera Aulagnier. Aulagnier, who married Castoriadis in 1968, was analyzed by Lacan between 1955 and 1961.97 She had been a strong advocate of Lacan during the International Psychoanalytic Association’s campaign against him, searching for a strategy of reconciliation and, when that failed, supporting the new Lacanian school. However, she and a number of colleagues soon became restive over what they perceived to be a growing authoritarianism within Lacan’s circle. When Lacan introduced the so-called “pass” in 1967, wherein analysts in training had to submit to an elaborate interrogation before a panel presided over by Lacan himself, Aulagnier and her allies fell into open opposition. In 1969 Aulagnier and several other disaffected Lacanians formed the Organisation psychanalytique de langue Française, otherwise known as the Quatrième Groupe. In reaction against Lacan’s authoritarian dictates, the group claimed a minimal ground of shared convictions, but above all it championed a pluralistic approach to theory and practice as well as an open institutional structure. Though Castoriadis was never a central figure in the group, his influence is manifest in its history. For example, a 1977 program urged its members to read his essay “Psychoanalysis, project and elucidation” on the subject of institution.⁹⁸ In turn, the circle of apostates from the École Freudienne clearly helped to shape Castoriadis’s animus toward Lacan and the refinement of his theory of the imagination. His first extended discussion of psychoanalysis, the 1968 essay “Epilogomena to a Theory of the Soul Which Has Been Presented as a Science,” was published just before the conflict produced an open schism in the École Freudienne. And for years there was a fruitful exchange of ideas between Castoriadis and Aulagnier; even after their marriage ended in a rancorous divorce, Castoriadis readily acknowledged his debt to her.⁹⁹

In numerous works beginning with “Epilegomena to a Theory of the Soul,” Castoriadis articulated the crucial difference between his and Lacan’s conceptions of the imaginary, and the difference bore directly on questions of the human subject, creativity, agency, and self-transformation. Although Castoriadis and Lacan both endorsed the “decentering” of the ego implied by Freud’s “discovery” of the unconscious, and in this sense both opposed American ego psychology, they did so in radically different ways. Lacan formed his core ideas on the ego in the 1930s by combining the work of the older French psychoanalyst Henri Wallon on the role of mirroring in the formation of the infant’s sense of identity with Alexandre Kojève’s Hegelian ideas about the role of a dialectical struggle for recognition in the development of subjectivity100 In his famous account of the mirror stage, Lacan depicted the ego as an imaginary construct produced through specular relations. The human infant overcomes its “motor incapacity and nursling dependence” by identifying itself with the image of corporeal wholeness presented by the “mirror,” both the literal reflection and the constitutive gaze of others. The fixation of the psyche upon the image of bodily wholeness “situates the agency known as the ego, prior to its social determination, in a fictional direction.”¹⁰¹ The reflected image of its own body gives the infant its first object of desire, but in the form of an alienating desire for the other, desire for what is missing, for the lack which structures the unconscious. The mirror stage thus sets into motion a libidinal dynamism that operates within the field of tension between fantasies of mastery and/or identification with the object of desire and the impossibility of such union or control. “In the order of the imaginary,” Lacan wrote, “alienation is constitutive. Alienation is the imaginary as such.”¹⁰² The “deflection of the specular I into the social I,” that is, the transition into the Symbolic Order, demands that the hold of the imaginary be relaxed; but, as Richard Boothby notes, “Even under the influence of the most farreaching effects of maturation and sublimation, the psychic organization remains at least partially oriented by the structure of the ego and can therefore never fully escape the orbit of the imaginary.”¹⁰³

As Castoriadis underscored in 1986, “We are speaking here, of course, of the radical imagination: not the capacity to have ‘images’ (or be seen) in a ‘mirror’ but the capacity to posit that which is not, to see in something that which is not there.”¹⁰⁶ In thus opposing Lacan’s theory of specularity, Castoriadis pursued a strategy uncannily similar to Fichte’s attempt to get beyond K. L. Reinhold’s belief that the self first comes to know itself through conscious reflection. Recalling our earlier discussion of Fichte’s reasoning, if the self is to be recognized in the mirror it holds up, it must first know who it is.¹⁰⁷ In similar manner to Fichte’s shift from viewing self-consciousness as a fact (Tatsache) to viewing it as an act (Tathandlung), Castoriadis hypothesized that “psychical life can exist only if the psyche is this original capacity to make representations arise, and, ‘at the start,’ a ‘first’ representation which must, in a certain manner, contain within itself the possibility of organizing all representations.”¹⁰⁸

Although Castoriadis’s commitment to a realist moment leads him to emphasize that the psyche, like society, leans on the “real,” he insisted that the originary capacity to give psychical representation to the drives—as well as the perceptual flux of representations, affects, and desires—cannot be traced exhaustively to the “real.” Castoriadis elaborated this claim for the rupture between the psyche and the real by introducing the idea of the “psychical monad.” In a description resembling Freud’s account of “primary narcissism,” the monad in its primal phase is enclosed in a self-referring circuit, dominated by the pleasure principle, not distinguishing between represented desire and satisfaction.109 The monadic core begins to break up under the pressure of bodily needs and the presence of another human being (in the typical case the mother). This socialization of the psyche involves “essentially imposing separation on it. For the psychical monad, this amounts to a violent break, forced by its ‘relation’ to others, more precisely, by the invasion of others as others, by means of which a ‘reality’ is constituted for the subject.”¹¹⁰ Even after the infant has become a “social individual” for whom social meanings have replaced purely private ones, the initial monadic pole continues to exert a powerful “tendency towards unification” over the psyche’s representations.¹¹¹ The psyche, Castoriadis wrote, is its own lost unity, and unconscious fantasy interminably reconstitutes this initial world “if not in its now inaccessible untouched unity, at least in its characteristics of closure, mastery, simultaneity and the absolute congruence between intention, representation, and affect.”¹¹² In the more socialized, conscious layers of the psyche, the unifying drive of the psychical monad operates as a power of synthesis creating the “relative unity of experience.”¹¹³

The theory of the monadic core served various functions for Castoriadis, but one of the most vital was surely that it gave him a better way to theorize the relationship between the imaginary and rationality. In his 1964–65 formulation, he sometimes seemed close to drawing a division between reason and the imaginary. Yet toward the close of “Marxism and Revolutionary Theory,” he made a highly suggestive claim regarding the beliefs of “ancient or archaic society”: “This imaginary does not merely hold the function of the rational, it is already a form of the latter and contains it in an initial and infinitely fertile indistinction; in it, one can start to distinguish the elements presupposed by our own rationality.”¹¹⁴ He seemed to have in mind something like the primitive classification that Durkheim and Mauss had described as a propaedeutic to our own activity of rational ordering. Through the theory of the psychical monad, Castoriadis could shed the lingering implications of an evolutionary development and instead theorize a dialectical relationship between rationality and the imaginary in the individual subject. Thus the same processes that could yield the madness of the psychotic could also yield the creations of reason:

Castoriadis maintained that the premise of an originary capacity that links drives to psychical representations is inherently necessary in the Freudian problematic, but Freud’s failure to make it explicit reflects a deeper reluctance to thematize the imagination as such. Although Castoriadis credited Freud for articulating the dynamics of the unconscious and posing the crucial question of how the initial psyche makes the transition from total self-absorption to socialized “normalcy,” he criticized Freud for pressing his discovery into the positivist mold of the late nineteenth century and he vigorously attacked later ego psychology for its attempt to domesticate the “scandal of the Unconscious.” This led Castoriadis to praise Lacan for shaking the psychoanalytic establishment’s ossified “positivist” self-understanding. Significantly, however, Castoriadis argued that Lacan reached an impasse precisely because he continued to adhere to a traditional scientific understanding of psychoanalysis. Hence, Castoriadis wrote that the Lacanians could not cope with the tension between the irreducibly singular individual, who is psychoanalysis’s object, and the general concepts and categories that analysis must necessarily deploy. Defeated by this tension, Lacanians conceive psychoanalysis only as a “projection, phantasm, delirium”—in other words, a defensive formation of the imaginary—insofar as it takes seriously the individual or as a “science” that establishes its rigor by exposing the individual as a product of the play of elements within a structure.¹¹⁷

If Castoriadis’s theory of the imaginary inverts Lacan’s, the same is true of his account of the socialization of the psyche and the torsions produced by that process. For Lacan, socialization is the (always incomplete) transition from the imaginary to the symbolic; for Castoriadis, it is the (always incomplete) entry of the individual psyche into a shared domain of social imaginary significations. Writing of Lacan’s conception of the subject, Malcolm Bowie observes, “the subject is no thing at all and can be grasped only as a set of tensions, or mutations, or dialectical upheavals within a continuous, intentional, future-directed process.”118 In opposition to this perpetual coming-into-being, the ego is the sum of imaginary resistances and misrecognitions that attempt to stabilize and unify identity. This conception of the tension between the subject and the ego points to a contrast fundamental to Lacan, which Bowie puts succinctly: “Where the [Symbolic] is characterized by difference, disjunction, and displacement, the [Imaginary] is a seeking for identity or resemblance. … Wherever a false identification is to be found—within the subject, or between one subject and another, or between subject and thing—there the Imaginary holds sway.”¹¹⁹ Where Lacan proposed a discontinuity between the imaginary and the symbolic, Castoriadis deepened the view, already articulated in his 1964–65 critique of Lévi-Strauss, that the imaginary and the symbolic are entwined, though not identical, as are the psychical imagination and the social-historical imagination.

This image of an irreducible gap between the psychical and the social-historical led Habermas to complain that Castoriadis’s theory of the psychical monad establishes a “metaphysical opposition” between psyche and society, which thereby fails to “provide us with a figure for the mediation between the individual and society.”120 Castoriadis countered this charge aggressively: “One should ask oneself, rather, what metaphysics is hidden behind the idea that every affirmation of irreducibility is ‘metaphysical.’ The answer is obvious: a unitary and reductionist metaphysics.”¹²¹ In fact, however, Castoriadis offered a more nuanced articulation of the relationship between psyche and society than he himself sometimes suggested. “There is no ‘human individual,’” he wrote. “There is a psyche that is socialized, and in this socialization, in the final result, there is almost nothing individual in the true sense of the term.”¹²² Consistent with the vision of the social-historical world as a sui generis domain he had formulated in the late 1950s, he insisted that socialization involves the psyche’s access to and accession to a “mode of being which the psyche could never give rise to starting from itself.”¹²³ This is a “violent” process directed at the “internally wish-generated stream of representations and affects.”¹²⁴ This language of violent imposition has misled some commentators, notably Peter Dews, to see between the social order of imaginary significations and the flux of representations that form the core of the individual psyche a permanent conflict that seems at odds with Castoriadis’s political optimism.¹²⁵

It is important to see that this imposition is also an introjection. For the psychical monad encounters not only obstacles and prohibitions to its own fantasy but also new “organizational schemata” that are enlisted in the psyche’s activity. This is the double message suggested by Castoriadis’s version of “sublimation”; he defined it as the “process by means of which the psyche is forced to replace its ‘own’ or ‘private’ objects of cathexis (including its own ‘image’ for itself), with objects that exist and are valid in and through their social institution, and out of these create for itself ‘causes,’ ‘means,’ or ‘supports’ of pleasure.” This process implies the “psyche itself as imagination, namely as the possibility of positing this for that,” even as it also implies “the social-historical as the social imaginary, namely as the positing, in and through the institution, of forms and significations which the psyche as such is absolutely incapable of bringing into existence.”¹²⁶ Psychical processes are codetermined by the social, yet the psyche’s entry into and participation in the social, in “language” and “doing as social activity,” does not erase the monadic core of the psyche, but satisfies or does not satisfy psychical needs, as the case may be. In place of Habermas’s desired “mediation” and in place of a “metaphysical opposition” between psyche and society, we find once again Castoriadis’s image of a nonempty intersection.

The opposition between Habermas and Castoriadis returns us to Lacan; as numerous scholars have noted, for all their differences, both Lacan and Habermas converge in their “linguistification” of the unconscious.127 Lacan’s famous claim that the “Unconscious is structured like a language” links it to the constituting order of the symbolic. Of course, this claim is notoriously enigmatic. It could mean that the unconscious has a structure like language, in which case the unconscious would itself be a closed order operating according to laws of combination of differential elements. Or it might mean that the symbolic allows the articulation of the unconscious through a kind of translation. As Lacan wrote, “Like the unnatural figures of the boat on the roof, or the man with a comma for a head, which are expressly mentioned by Freud, dream images are to be taken only on the basis of their value as signifiers, that is, only insofar as they allow us to spell out the ‘proverb’ presented by the oneiric rebus.”¹²⁸ Or, it could mean, as Boothby suggests, that the symbolic volatilizes the imaginary, confronting the fixations of the ego with the displacements and discontinuities of language. Against this penetration of the unconscious by the symbolic, Castoriadis insisted on what could be described as a more orthodox Freudian division between thing representations and word representations, between primary and secondary processes. In the unconscious, he argued, “there is no representation of words as words that would convey some sort of rationality; there is not, and cannot be, any symbolism, anything symbolic.”¹²⁹

In insisting so categorically upon a nonlinguistic and hence nonsocial psychical core, Castoriadis adopted quite an extreme position vis-à-vis a problem that became particularly pressing to many on the French left after the disappointment of May 1968, namely, how society penetrates and reproduces a socially normative individual.130 Sherry Turkle writes that whereas many on the left prior to 1968 had still denigrated structuralism as overly deterministic, May 1968 lent credence to structuralism’s claim that man is inhabited by the signifier rather than freely creating it: “What had seemed reactionary in structuralism now seemed merely realistic.”¹³¹ Lacan’s influence found its way into the women’s movement (Antionette Fouque, Luce Irigaray, Michèle Montrelay), French antipsychiatry (Félix Guattari, Gilles Deleuze), French Maoism (Jacques-Alain Miller, Judith Miller, Lacan’s daughter), as well as textual theory (Philippe Sollers, Julia Kristeva).¹³² Yet, if Lacan’s depiction of a symbolic network enveloping the individual thus reshaped the sensibility of many left-leaning French intellectuals after 1968, the abiding leftist investment in progressive change produced a certain kind of schizophrenia: Precisely the omnipresence of the symbolic order prompted ideas of total escape, a leap into another way of being and speaking. Given what the failure of 1968 seemed to show, that the powers of the symbolic order could recuperate and reabsorb even the most radical of emancipatory impulses, the idea of revolutionary change had to migrate to a neverland bounded by its own pristine impossibility. This hardened into an impasse what Peter Starr defines as Lacan’s “tragicomic” politics, which occupied a “vacillatory” space “where the prospect of a partial release from self-division intersects with the full awareness of all that opposes human redemption.”¹³³ Lacan was drawn to the left out of respect for the truth of revolutionary desire, yet he criticized emancipatory ideology for its pursuit of an impossible happiness; and he extended his criticism of the demand for the unity of the “One” (truth, meaning, system, ego) to political revolution. In this pursuit of an imaginary wholeness, he saw only the inevitable repetition, the specular doubling, of power’s own monological impulse to unify. He saw in every transgression in the direction of jouissance an enforcement of the Law that opposes, yet defines, transgression.¹³⁴

Castoriadis’s writings against the avant-garde styles of thought in the 1970s recognized clearly what Peter Starr later described. For example, in a 1977 piece titled “The Diversionists,” Castoriadis argued that structuralism arose as an evacuation of “living history” and political activism. It could hardly be more opposed to the forces that erupted in 1968, he insisted. The waves of French theory that commanded the stage in the 1970s completely sidestepped what Castoriadis regarded as the main challenge posed by the events of 1968, namely how this explosion of political activism might go beyond its initial stage without losing its creativity, how this “fantastic deployment of autonomous activity” might be able to “institute lasting collective organizations that express it without drying it up or confiscating it.”¹³⁵ Castoriadis formulated this position even more forcefully in a 1986 critique of Alain Renaut and Luc Ferry’s book La Pensée 68. “It is strange,” he wrote,

Far from spurring revolt, he argued, the French ideology provided legitimation for “withdrawal, renunciation, noncommitment, or … a punctilious and measured commitment.”¹³⁷

Castoriadis also saw clearly the impasse structure that Peter Starr describes as one of the key features of 1970s French thought. Indeed, it is stunning that Castoriadis recognized this within weeks of May 1968 in his contribution to La Brèche, the volume he co-authored with Claude Lefort and Edgar Morin.¹³⁸ As he conceded, the upheavals of May speak to the dimension of failure in modern radical movements, the “enormous difficulty involved in extending critique of the existing order in practical and positive ways, the impossibility of assuming the goal of an autonomy that is at once individual and social by establishing collective self-governance.” This failure has haunted modern revolutions; but Castoriadis cautioned that the failure has rarely been total. Like the Paris Commune, he predicted that May ’68 is likely to leave deep marks on French mentalities and patterns of political behavior. And then, in an observation that anticipates the structure in 1970s thought described by Peter Starr, Castoriadis detected a tendency to overlook this historical remainder because the modern political imaginary is wedded to antinomic models of power and politics that fantasize either a leap into an unheralded kingdom of freedom or a fall back into the oppressive structures of the past.¹³⁹

Aimed at autonomy, psychoanalysis should take the “subject” not as its point of departure but as its end goal, as its project. Of course, we have seen that Castoriadis recognized that socialization intrudes on the psychical monad to produce an individual “that functions adequately for itself most of the time … and, above all, functions adequately from the point of view of society.”¹⁴¹ Yet he chided his contemporaries because “an enormous part of the rhetoric of the sixties and seventies concerning the subject as a simple effect of language and its ‘unbeing’ was in fact questioning only this social individual.”¹⁴² In contrast, Castoriadis insisted on the capacity for reflective self-representation and deliberate activity that allows the “subject or human subjectivity properly speaking” to put social boundaries and even itself into question. This type of subjectivity is an ongoing project, neither a static reality, nor an unattainable ideal, nor a fiction, nor a guaranteed telos. Situating the question of subjectivity and autonomy in this way makes it imperative that history be thought in its openness to the emergence of radically novel innovations. In this regard, Lacan, in common with all structuralists, excluded the essential dimension: temporality. In a vital passage, Castoriadis wrote of the meaning of temporality in the history of the psyche and in the practico-poietic activity of psychoanalysis:

Castoriadis’s recovery of the temporal dimension, of history in the strong sense as radical alterity—“ontological creation deploying itself as history”—united his engagement with psychoanalysis with the redefinition of radical politics he had first offered in his writings of the early 1960s. The “project of autonomy” animated Castoriadis’s efforts to rethink the social-historical world as the domain of human creation, to free it from deterministic or functionalist logics, yet to understand its relationship to a “real” world that permits this creation. For Castoriadis, it is the abyss of imagination that constitutes the democratic opening by bringing forward “new figures of the thinkable.” It is the radical imaginary that opens the possibility of a struggle for a new “relation of society to its institutions, for the instauration of a state of affairs in which man as a social being is able and willing to regard the institutions that rule his life as his own collective creations, and hence is able and willing to transform them each time he has the need or the desire.”¹⁴⁴ In this vision of radical democracy one readily discerns traces of a classic theme of German political idealism, the association of freedom with the emergence of a self-conscious recognition that it is human will that creates the institutions of society. Yet if Castoriadis drew from that tradition, he recast the struggle for a self-conscious relationship between citizens and their institutions in terms that rejected the ideals of social transparency, determinism, and every sort of absolute. Politics, like the individual psyche, is part of a world that is obscure and unmasterable, yet capable of elucidation. Consistent with his rejection of foundationalist logics, Castoriadis maintained that the value of autonomy cannot be “grounded” or “proven”; at most autonomy can be “reasonably argued for and argued about” once it has emerged historically in the social imaginary^.145

The group Socialisme ou Barbarie disbanded once and for all in 1967. That marked the end of Castoriadis’s direct involvement in a militant group. He had, in any case, already made a decisive move from the marginalized intellectual and political milieu of the extreme left to a much broader and explicit engagement with the intellectual currents of the period when he began to publish “Marxism and Revolutionary Theory” in serialized form in Socialisme ou Barbarie. This trajectory reached its apogee with the later sections of The Imaginary Institution of Society, composed in the early 1970s, which articulated all the themes that would preoccupy Castoriadis for the rest of his career. The shift in style and content in the years since roughly 1960 was mirrored by a shift in the loci of Castoriadis’s activities: he became a practicing psychoanalyst in the early 1970s and he was elected a directeur d’études at the École des hautes études en sciences sociales in 1980. A Greek expatriate who bypassed French education on his way to the radical peripheries of French politics, Castoriadis arrived at the upper reaches of French academe along a highly unconventional path. He always remained something of an outsider within the cliquish intellectual life of Paris, which he was sometimes able to regard with ethnographic curiosity, that is, when he was not attacking it with fury. His outsider status may help to explain his resistance to successive waves of intellectual fads, from communist fellow-traveling, Sartrean existentialism, and structuralism to poststructuralism, Foucauldianism, the new philosophers, and so on. Yet that explanation would ascribe too much causal power to sociological position, whereas what really seemed to sustain him was a clear-sighted fidelity to his own political commitments. With remarkable consistency, from the earliest days of Socialisme ou Barbarie onward, Castoriadis championed the idea that, in myriad ways, people may act creatively to alter the conditions of their lives. This idea not only weathered the changes in Castoriadis’s writing and spheres of action, but it gained in power as Castoriadis articulated a deepening critique of Western ideas about being, society, history, and selfhood.

Castoriadis liked to insist upon the reality of creatio ex nihilo in the hope that this most extreme image of creation would sweep away “the subterfuges and sophisms concerning the question of the new: either there is creation or else the history of being (therefore also of humanity) is interminable repetition (or eternal return).”147 Yet he was invariably quick to add that creation is ex nihilo, but not in nihilo or cum nihilo. That is, even if new forms cannot be deduced from existing elements, they do not emerge in a vacuum nor are they “made” from nothing. The history of Castoriadis’s own thought certainly supports this image of creativity. As we have seen, his break from Marxism and attempt to rethink radical autonomy unfolded through a deep and evolving engagement with the resources of modern philosophy, from German idealism to phenomenology, and the psychoanalytic tradition. A crucial dimension of this process was his polemically charged relation to competing currents, from opposition to Stalinism and Trotskyism in the name of true socialism, to opposition to Marxism in the name of revolution, to opposition to structuralism in the name of human agency.

Castoriadis’s argument with structuralism, so central to the years of his most fruitful theoretical development, in turn evolved from a critique of Lévi-Strauss’s exclusion of the imaginary from the symbolic to a critique of Lacan’s attempt to reduce the imaginary to a structure of capture and fixation. In the process, Castoriadis went from a superficial but essentially positive to a deep but explicitly hostile view of Lacan. Seen from a different angle, however, it is striking that Castoriadis always treated Lacan exclusively as one of the high priests of structuralism. Once he arrived at this negative judgment, he never acknowledged tensions, ambiguities, or developments in Lacan’s position. It must be said that a lack of nuance or generosity toward his opponents was typical of Castoriadis, for whom the political stakes of theoretical debate were always immediate and paramount.¹⁴⁸ Undoubtedly, his work represents a vital and original contribution to the thought of the late twentieth century across a spectrum of fields and questions, but his approach to the work of other thinkers was almost invariably polemical. So, for example, his erstwhile comrade Claude Lefort once noted that Castoriadis’s “desire to desanctify Marx, to shatter the myth attached to his name, while certainly legitimate, leads him to over-emphasize his rupture with Marx.”¹⁴⁹ For similar reasons, Castoriadis may have overemphasized his rupture with the structuralism of Lévi-Strauss and Lacan, although, as this chapter has shown, far from discrediting his innovations, links to the styles of thought against which he rebelled lend his work additional power and interest.

The distance separating Castoriadis from Lacan was substantial, but not absolute; and this becomes even clearer if we recall Castoriadis’s use of Fichte’s idea of Anstoß, the check or stimulus that human significations experience when they encounter the inexhaustible otherness of the thing-in-itself. Castoriadis meant this to shatter structuralism’s closed order of language and law. Yet we must not forget, as Slavoj Žižek repeatedly reminds us, that Lacan himself turned away from the structuralism of the 1950s, when the symbolic order functioned as a “proto-transcendental structural a priori,” which “in advance predetermines the subject’s acts, so that we even do not speak, but ‘are spoken’ by the Other.”150 The later Lacan turned increasingly to the real, that kernel or remainder of the unconscious that can never be dissolved or integrated into the symbolic order. It is as the philosopher of the real, of the disruption of the symbolic order, and hence, of the flawed symbolic, that Lacan has most palpably influenced the discourse of post-Marxism. Slavoj Žižek stakes his entire oeuvre upon the contrast between the early and the later Lacan, but even Claude Lefort and Marcel Gauchet, while certainly not disciples of Lacan, nonetheless drew upon him. If Castoriadis revealed the limits of Lacan’s structuralist paradigm for a theory of radical politics, other thinkers perceived in the Lacanian theory of the real a more promising pathway for political thought.