Among the concepts proposed by Althusser in the course of his famous symptomatic reading of Marx’s Capital, structural causality plays a central role. Extrapolated from Marx’s writings via a detour through the philosophy of Spinoza, it came to represent the concept in which Althusser summed up ‘Marx’s immense theoretical revolution’.  As is well known, for Althusser, by breaking with the other two models of causality available in Marx’s time (the mechanistic causality derived from Descartes and the ‘expressive’ causality derived from Leibniz and above all Hegel), ‘structural causality’ made it possible for the first time to think of history as a process deprived of any essence and telos, without subject and without end – or, as Althusser puts it in another formulation, as a ‘structure of structures’ without any centre.  It was, therefore, the concept that condensed the antihumanist reading produced in the 1960s by Althusser and his collaborators, one that soon became famous under the label ‘structural Marxism’.
If the centrality of structural causality to the overall project of recasting Marxism is beyond doubt, the fate of this concept is curious. For one thing, the number of pages devoted to the explicit theoretical elaboration of structural causality in Reading Capital is rather few (fourteen in the French edition). More importantly, the concept will soon disappear from Althusser’s discourse, in his attempt to correct his so-called ‘theoreticism’ and to reject the allegations of ‘structuralism’, to which ‘structural causality’ evidently gave rise through its name. It is well known that structural causality was one of the concepts that led to fierce criticism of Althusser. For his proximity to structuralism, he was accused of denying history; for his reliance on Spinoza, he was accused of endorsing a metaphysical concept of necessity that did not leave any space for freedom or – more importantly for the purposes of this article – for contingency. Famously (to limit ourselves to the anglophone reception), Althusser was attacked by the British Marxist E.P. Thompson, who accused him of endorsing a deterministic philosophy of history that asserted the timeless reproduction of the structure of a certain mode of production.  The same type of critique was put forward, first, by Hindess and Hirst, and then by Laclau and Mouffe, in their Hegemony and Socialist Strategy. There, the latter authors argued that the most vital part of Althusser’s project is to be found in his concept of ‘overdetermination’, but that soon after the essay (dating from 1962) in which that concept was propounded, Althusser retreated towards a hopeless metaphysical rationalism that culminated in the concept of structural causality. In their reading, Althusser’s attempt to pass from Hegel to Spinoza in his reading of Marx only resulted in the substitution of a strictly logico-mathematical necessity for a teleological necessity governing history.  In his now-classical study on Western Marxism, Perry Anderson remarked that Althusser’s Spinozism was so deep that the metaphysical determinism of Spinoza could be found without any modification, in particular in the ‘implacable logic’, as he called it, of ‘structural causality’.  As Peter Thomas has recently noted, Anderson’s study of the relationship between Althusser and Spinoza was seminal: most later interpretations of Althusser’s Marxism merely repeating Anderson’s reading without any modification. 
In this article, I argue that Althusser’s concept of structural causality cannot be reduced to a metaphysical necessitarianism of Spinoza’s kind or a structuralist determinism. My aim is to show that, far from being a strictly determinist concept, structural causality was the concept through which Althusser attempted for the first time to develop a logic capable of including contingency as a structural dimension, and that, far from asserting the timeless reproduction of the structure (i.e. of a certain mode of production), it should be read as the concept through which Althusser tried to propose a non-dialectical theory of structural change.
Now, it is well known today, after the publication of Althusser’s late writings, that he proposed in the 1980s a new philosophy for Marxism that he named materialism of ‘the encounter’ or of ‘the swerve’.  However, I do not intend to project the problematic of the late Althusser onto his theories of the 1960s. I will take an opposite approach and attempt to show that ‘structural causality’ was, in fact, the site of Althusser’s effort to flesh out a new way of thinking about the interrelation of necessity and contingency in history, rooted in a long-standing preoccupation with the notion of ‘necessity’ that can be dated as far back as the late 1940s. In order to do so, I will proceed via a reading of some crucial moments of the early Althusser (that is, Althusser before the 1960s), so as to show his attempt both to deconstruct and to reconstruct the notions of necessity that he found in Hegel and the orthodox Marxism of the 1950s. I will then move on to a ‘symptomatic reading’ of structural causality itself as conceptualized in Reading Capital and as developed, in the months after the publication of the book, in some posthumously published texts.
Unlike the work produced by Althusser after 1980, the writings dating from before the 1960s have received very little attention. However, an overview of the most important writings preceding the publication of For Marx and Reading Capital is highly instructive. One of the main lines that emerge is one concerning the status of ‘necessity’ within Marxism, a circumstance that allows us to argue that the knot of problems centred around the relationship between ‘causality’ and ‘necessity’ in history forms the core of Althusser’s philosophical preoccupations in the run-up to his reading of Marx in the 1960s. The first time that Althusser tackled the problem is as early as 1948 in his dissertation on Hegel.  Although this dissertation has been considered (and quickly dismissed) as proof that – before becoming fiercely anti-Hegelian – Althusser had indeed been Hegelian,  a careful reading suggests that Althusser was already pointing out the problems of a humanist reading of Marx and asserting a sharp distinction between the young Marx and the Marx of Capital. He did so by focusing on Marx’s relations to Hegel’s philosophy of history, thus anticipating the later critique that he will level against Hegel in the 1960s. What is characteristic of this early text is that the central question of the Hegel–Marx relationship is not centred around the problem of science and ideology (as it was to be in the 1960s), but bears instead on the question of the status of necessity. In his reading of Hegel (which demonstrates, contrary to what has sometimes been suggested, that Althusser knew the entire corpus of Hegel’s writings perfectly), Althusser is particularly attentive to the structure of the central operator of Hegel’s philosophy: his concept of ‘concept’. Althusser describes it as ‘pure interiority’ with ‘no outside’ (later on, this will be the very definition of ideology), as a ‘process of envelopment’ whose activity of interiorization aims to posit itself as ‘the origin of the origin’ and to recover all ‘free contingent events’ as ‘moments of the fully accomplished totality’.  But his attention goes in particular to the fundamental consequences that such a view has for the conception of historical necessity. In Hegel’s hands, so argues Althusser, historical necessity is merely the (teleo)logical necessity of the concept rendered indifferent to the concrete determinations of history, which in turn are endowed only with an apparent substantiality incapable of ‘affecting’ the unfolding of historical necessity itself.  What is interesting for our purposes here is not the correctness or otherwise of such critique, but rather the consequences that Althusser draws for his reading of Marx, to whom he turns in the last sections of this text. The question that Althusser poses here is whether or not Marx’s critique of Hegel is a consistent one. Althusser argues that Marx’s explicit criticism of Hegel (present in his early writings) is still trapped in a Hegelian framework. Assuredly, as Althusser notes, Marx refuses to see in the Prussian state the ‘end of history’. Nevertheless, such a rejection takes the form of a mere displacement of the idea of an accomplishment of history towards the future. For all Marx’s criticism, in his analyses of the vicissitudes of ‘alienation’ we ‘find Hegelian necessity again, in its most rigorous form – that of the concept’.  However, the real innovation is to be found, according to Althusser, in Capital. For Althusser here, Marx understood thatthe transcendental was history, but he did not consider it possible to think history in general, apart from the concrete content of the dominant historical totality … he did not posit the categorial totality as eternal. 
In his later work, argues Althusser, Marx abandoned the idea of the ‘end of history’ as well as the idea of history as a teleological process, paving the way for a new conception of the transcendental and therefore of historical causality:
if we abandon the idea of the end of history and the eternal nature of meanings [l’eternité des significations] … then history becomes the general element in which we move and leave … the concrete transcendental … that conditions and determines us. But since history is not over, there is no eternal transcendental logic, but rather, at every instant, an articulated historical structure which dominates the world in the manner of an a-priori, and conditions it. The reality of history resides, from this standpoint, in the dialectical nature of the structure that conditions events, but is also transformed by them in its turn. The historical totality is a concrete, dialectical transcendental, a condition modified by what it conditions. 
What is important here is that for Althusser, while Capital provides us with a new ‘transcendental analytic’, the transcendental itself is not the Träger of any metaphysical necessity, but only of what he calls a ‘de facto necessity’.  This perspective, which aims to abolish the ontological separation between the transcendental and the empirical (and which could be called ‘empirical transcendentalism’), implies a new, ‘weaker’ conception of necessity bound up with an ‘articulated historical structure’ that both conditions and is conditioned by the ‘manifold that it conditioned’.  The dissertation ends precisely on this problem of an adequate conceptualization of ‘necessity’ in Marxism, which Althusser resolves through the concepts of ‘de facto necessity’ and ‘empirical transcendentalism’, although they are left largely undeveloped. However, this attests that the concept of necessity was very much at the core of his preoccupations and that it was also the central concern, from as early as 1948, of his relationship with Marxism. Furthermore, it shows that it was by reflecting on this question that Althusser introduced, well before For Marx, the idea of a break between the early Marx and the Marx of Capital.
The same concerns with the category of necessity in Marx and Marxism are present in at least three other writings from the 1950s. In the course of the lectures on the philosophy of history that he gave during the 1950s at the ENS, for example, Althusser grapples with the question of history as a closed or open process.  The central idea is, again, that what distinguishes Hegel from Marx is that Hegel can only think of history as a ‘closed’ process – that is, from the point of view of the Spirit reconciled with itself at the end of the historical unfolding. By contrast, Marx provided us with the means to conceive of history as an open process.  In this course, Althusser abandons the idea of an ‘empirical transcendentalism’ and shows a more rigid orthodoxy. For instance, when discussing the issue of the laws of history, he relies quite heavily not only on Lenin but also on Stalin, and states that ‘it is the functional nature of the relations of production that allow us to comprehend the necessity of the transition from determinate social conditions to different ones, i.e. revolutions’.  However, the course is particularly interesting because it shows the central tension in Althusser between the exigency of orthodoxy, leaning towards the ‘iron laws of history’, and the need to come up with a less simplistic conception of necessity than the one provided by the dogma.
The solution that Althusser introduced here is in more than one sense much less interesting than the one he outlined in his dissertation. In the attempts to limit the validity of the ‘laws’ discovered by Marxism, he appeals to an ‘inexhaustible reality’ that precedes the science of history and ‘always surpasses it’.  Such an appeal must be read as an attack on the idea of absolute knowledge; that is, as a passage levelled against the idea that Marxism is a philosophy of history stating laws that are valid once and for all. The obvious problem is, however, that such an appeal to the need of rectification of the outcomes of scientific research stands in contrast to, or at least in a problematic relationship with, the very idea of the ‘necessity of the transition’ that Althusser nonetheless maintains. If history is not reducible to a ‘law’, and if there is always a sort of ‘excess’ over the conceptualization of it, how can one argue that there is something like a ‘necessity of the transition’? This shows perhaps the difficulties that Althusser was facing in those years of almost strict orthodoxy. But it also shows that even in those years he was at odds with the Marxist (or Stalinist) conception of necessity, and somehow attempted to correct it though an operation of Ptolemization that, instead of calling into question the central tenets of a theory, was content to add epicycles in the hope of making things work.
However, the need to produce a new concept of causality is raised in an article written by Althusser in response to Ricœur’s and Aron’s attacks on the very idea of a science of history. Althusser responds to Ricœur’s argument by conceding that history may be about ‘an inexhaustible nature of man’, or about freedom, but refusing to accept their conclusion that such a view could be used to ‘refuse in advance the pretences of history to objectivity’.  The problem that Althusser has with Ricœur and Aron is crystallized in an interesting remark about historical causality. Althusser argues that the problems raised by the very opposition put forth by Ricœur must be resolved not by philosophically opposing two domains (histoire vécue against science), which do not belong to the same ‘level’, but through the elaboration of a concept of causality capable of rendering history intelligible;22 that is, on an epistemological level. Comparing this answer with the solution Althusser proposed in his course – which appealed, precisely, to the ‘inexhaustible’ character of history – it is difficult not to notice that the argument is the same, even if Althusser does not use it as a means to deny to history the status of science. This shows that, in all probability, at that time Althusser was not insensitive to the critiques of Marxism launched by such thinkers as Aron and Ricœur and attempted to attenuate the Marxist concept of necessity, or at least to problematize it. 
A further step – and indeed a decisive one – in this direction is taken by Althusser in his book on Montesquieu, written a few years later – crucially, after 1956.24 In this work, which shows a degree of independent thinking comparable to the dissertation on Hegel (to which in a sense it returns), Montesquieu is seen as a direct precursor to Marx in that ‘he was the first person before Marx who undertook to think history without attributing it to an end, without projecting the consciousness of men and their hopes onto the time of history.’  More importantly, Montesquieu became for Althusser a means to develop a new schema of historical causality based on the idea of ‘multiple determinations’. In this sense, Althusser’s reading of the Ésprit des lois can be seen as the attempt to develop a new concept of complex necessity.
The first point that Althusser makes is that Montesquieu provides a ‘dynamic’ model for the conceptualization of historical totality based on the dialectics between two elements: the ‘nature’ of the government (the question of who holds power) and the ‘principle’ (the passion by which a certain government is made to act).  It is the relationship between these two elements that allows Montesquieu to posit the problem of the ‘motor of history’. Arguing against the idea that Montesquieu theorized a perfect circularity of historical totalities – a position held at the time by Cassirer – Althusser insists that between ‘principle’ and ‘nature’ there exists, in Montesquieu, an asymmetry, which posits the dominance of one element over the other: the dominance of the ‘principle’. Althusser argues that (1) this dominance performs a distribution of efficacy: the principle is dominant, but ‘nature’ can act back on it; (2) it breaks with circularity, with the idea of an ‘expressive totality’, according to which, as Althusser will say later on in Reading Capital, all the elements of a certain totality are directly determined by a single principle, thus forming a whole with no internal differentiation. However – and it is here that we approach the question of a complex necessity – Althusser argues that the internal asymmetry is, in turn, sustained by another type of causality, which Althusser extracts from a reading of the second part of the Ésprit des lois. In the second part of his book, Montesquieu adds, in the form of empirical observations, a series of determining factors other than the two previously mentioned, such as climate, soil, religion and others. As Althusser notes: ‘in front of the new determinant factors suggested [by Montesquieu] … it is hard to avoid the impression of disorder. The unity of a profound law has turned into a plurality of causes. The totality is lost in a list.’  In reality, this list is only apparently one. For Althusser, the impression of disorder is due to the fact that Montesquieu did not produce the concept of their causality. For Althusser, what characterizes them is that their efficacy is not a direct one: these factors act on the ‘principle’ – on the dominant element of the dialectics of history – only through an ‘indirect causality’ that breaks with a mechanical type of determination.  This means that the ‘principle’, the dominant or determinant factor in Montesquieu’s model, is in itself determined by factors that are not separable from it. Althusser criticizes Montesquieu for his failure to conceptualize this idea and, to account for it, he introduces the notion of a causality through ‘conjunction’ and ‘encounter’.  just when they are acting on the government and determining certain of its essential laws, all these causes apparently so radically disparate, converge on a common point: the customs, morals and manners of being, feeling and acting that they confer on the men who live within their empire. From their encounter [rencontre] arises what Montesquieu calls the spirit of a nation. 
Thus, the very dialectics between principle and nature is sustained by another causality that encompasses a whole set of heterogeneous factors. The concept of ‘encounter’, introduced here for the first time, will soon become central to Althusser’s attempt to come up with a notion of causality capable of mediating or, better said, capable of articulating, necessity and contingency. The issue of contingency is certainly not present at this point, even if the idea of ‘encounter’ or ‘conjunction’ is introduced to displace the idea of a mechanical or expressive necessity commanding the unfolding of a dialectical totality. The concept of structural causality will inherit and radicalize these issues, introducing the question of the relationship between necessity and contingency to the very heart of Althusser’s recasting of Marxism.
: textual symptoms
As mentioned above, one of the criticisms that has been levelled against Althusser is that, with Reading Capital, he retreats from certain (good) intuitions that formed the core of For Marx – namely the concept of overdetermination – towards a (bad) metaphysical rationalism. Especially in the field of what is often referred to as post-Marxism, overdetermination is considered to open the way towards contingency, whereas ‘structural causality’ would close that door by asserting a strict determinism. However, the problem with the identification of ‘contingency’ as a central question for Reading Capital, and as a problem intimately related to the elaboration of structural causality, is that although a survey of all the occurrences of the word in the book reveals the persistence of its presence, it also reveals the unrigorous way in which it is used.  In many places, Althusser seems to reject it.  This can only reinforce the idea that necessity was for him the central question. However, a textual symptom, located in a strategic passage, suggests otherwise. In a passage from the introduction to Reading Capital (written after the seminar in which the papers collected in the book was completed) Althusser argues that, in the wake of the works of Foucault, Canguilhem and Bachelard, we are compelled to abandon the idea of history as a single uniform continuum, and that ‘we are beginning to conceive this history as a history punctuated by radical discontinuities’. Immediately afterwards, he adds:
We are thereby obliged to renounce every teleology of reason, and to conceive the historical relation between a result and its conditions of existence as a relation of production, and not of expression, and therefore as what, in a phrase that clashes with the classical system of categories and demands the replacement of those categories, we can call the necessity of its contingency. 
The expression ‘necessity of contingency’ is here mobilized against the idea of a simple and continuous teleology of reason, which Althusser (arguably a little hastily) attributes to the philosophy of Enlightenment and to Hegel.  It is likewise opposed to what we can call a ‘logic of expression’, for which Althusser wants to substitute a ‘logic of production’. This opposition is, indeed, revealing. Given that, in For Marx and especially in Reading Capital, Althusser criticizes the concept of ‘expressive causality’ for its reductionism of the elements of a determinate ‘whole’ to an inner essence or principle, arguing for the new type of causality represented by ‘structural causality’, we have here the indication that ‘structural causality’ occupies the conceptual space indicated by the phrases ‘necessity of contingency’ and ‘logic of production’. It is certainly strange that Althusser does not bring structural causality itself into the discussion at this point. However, it is clear that we are confronted here with a decisive substitution: ‘necessity of contingency’ plays the role that will be taken on by structural causality, which will be opposed to the ‘expressive whole’ (that is, to a logic of expression) and to the Leibniz–Hegelian ‘expressive causality’.
Nonetheless, immediately after proposing this category of ‘necessity of contingency’, Althusser fails to flesh it out. What is worse, he never mentions it again in Reading Capital, not even when, in chapter IX, he develops the concept of structural causality. We are confronted, at a textual level, with a problem: the ‘enigma’ of the ‘necessity of contingency’ – this category of which we are only told that ‘it clashes with the classical system of categories’ – seems to direct us to the concept of ‘structural causality’ (to which the opposition to the ‘logic of expression’ points), and yet ‘structural causality’ undeniably relies on the Spinozist concept of immanent causality, and for Spinoza ‘in rerum natura nullum datur contingens’;35 in Althusser’s conceptualization of structural causality the problem of contingency is never mentioned per se. Notwithstanding this lexical problem – which I take here as a symptom – the opposition that we stressed suggests that structural causality is associated by Althusser himself with the category of ‘necessity of contingency’, and therefore, in some way or another, Althusser considers the issue of contingency as central to his attempt to recast Marxism on the basis of the category of structural causality.
Absence and presence of the structure
Following this textual symptom, the problem is to understand how and if Althusser, in his elaborations of structural causality in Reading Capital, effectively manages to incorporate the ‘necessity of contingency’, and where precisely this incorporation can be found. Here, a certain displacement is needed with respect to the usual way of reading the concept of structural causality, which (when it isn’t dismissed as simply metaphysical or deterministic) insists on a certain tension between a Lacanian idea of the ‘absent cause’ and a more Spinozist model of an immanent causality).
I want to suggest that structural causality is actually premissed upon a rejection of Lacan’s (structuralist) model of the absent cause, and also on a crucial modification of Spinozist ontology of immanence. In order to disentangle the presence of ‘necessity of contingency’ – or at least Althusser’s attempt to make structural causality function as a conceptual elaboration of this category – we need to move centre stage a category that is usually neglected in the analyses of structural causality: the category of ‘determinate absence’, and the reciprocity that this category entails between ‘overdetermination’ and ‘underdetermination’. This corresponds to the specific innovation introduced by Althusser in Reading Capital with respect to the idea of the overdetermined contradiction put forth a few years earlier. 
The rejection of the model of the ‘absent cause’ is evident in the way in which Althusser treats the concept of metonymic causality, a concept introduced by Jacques-Alain Miller that formalized for the first time, in the wake of Lacan, the idea of a relationship between causality and absence. In ‘Action of the Structure’ (1964), Miller presented this type of causality as a point of transition between Marxism and psychoanalysis, developing it into a more general theory of the dependency of the subject on the structure via the introduction of the idea of the ‘absent cause’. According to this model, following the perception of the structure by a subject, the structure governs the real by ‘not being there’.  If this theory was later used in Reading Capital, especially by Rancière in relation to fetishism, the concept of ‘structural causality’ as deployed by Althusser nonetheless stands at a certain distance from Miller’s specific attempt to make ‘metonymic causality’ function in both psychoanalysis and Marxism. Whereas Miller’s concept involved a perceiving subject in its relationship with a determinate structure, what is at the core of ‘structural causality’ seems to be hardly conceivable by means of ‘metonymic causality’: namely, the efficacy of a structure on elements that are not necessarily ‘perceiving subjects’, and above all the efficacy of a structure on subordinated structures (plural). 
Even though Althusser does not criticize the concept of metonymic causality explicitly, nonetheless he tries to subordinate the idea of absence that it implies to the idea of the presence of the structure in its effects; that is, to subordinate it to a Spinozist conception of immanence. For example, in a passage that was suppressed from the second edition of Reading Capital, Althusser points out that the concept of Darstellung (which, for him, was Marx’s name for structural causality, or at least the name by means of which Marx came closest to naming the concept of this new type of causality) can mean both the absence and the presence of the structure in its effects. But immediately afterwards he notes:
I believe that understood as the concept of the efficacy of an absent cause, this concept is perfectly fitted to designate the absence in person of the structure from its effects considered in the oblique perspective of their existence [perspective rasante] adding that it is necessary to insist on the other aspect, the immanence or presence of the cause in its effects. 
To be sure, this formulation (‘oblique perspective’) is quite obscure, and the fact that Althusser suppressed the passage in the second edition attests that he was aware of that. However, it is clear that Althusser intends to limit the idea of an absent cause to a certain perspective, and this move has the effect of limiting the validity of metonymic causality: the risk that Althusser sees in the idea of an absent cause is that it reintroduces surreptitiously an idea of transcendence by referring to an absence that would automatically be conceived as a ‘beyond’.  Nonetheless, the priority attributed by Althusser to the presence of the cause in its effects does not mean that within structural causality, absence does not play any role.
In the chapter on time in Reading Capital,  Althusser attempts to develop the concept of historical time according to the concept of structural causality, and to flesh out the concept of ‘conjuncture’. He does so by arguing that the Marxist conception of the ‘whole’ has a crucial consequence for the theorization of the temporality of this ‘whole’. As the levels or instances are not, as in the Hegelian model, reducible to an original simple unity, ‘each of the different levels of the whole does not have the same historical existence’, and it is thus necessary to ‘assign to each level a peculiar time, relatively autonomous and hence relatively independent, even in its dependence, of the “times” of the other levels’.  Thus, unlike the time proper to the expressive whole, this time cannot be subjected to an ‘essential section’, a neat cut that reveals the ‘coeval character’ of all the levels of the whole (which are not, in the expressive whole, properly speaking, different from each other). To account for the double situation of dependence and independence of the levels of the whole, Althusser introduces the notion of underdetermination:
to speak of differential historical temporality absolutely obliges us … to think, in its peculiar articulation, the function of such an element or such a level in the current configuration of the whole; to determine the relation of articulation of this element as a function of other elements, of this structure as a function of other structures; it obliges us to define what has been called its overdetermination or underdetermination, as a function of the structure of the determination of the whole. 
First, let us note that in this context, overand underdetermination are not used in reference to the concept of contradiction. Earlier on, in For Marx, Althusser had spoken of the overdetermination of the contradiction; later on, he would return to these terms saying that there exists a ‘threshold’ of determination of the contradiction that must be reached for a revolution to happen.  Here, though, the idea of a ‘threshold’ is not mentioned. Althusser uses overand underdetermination in reference to the levels in a general situation. In this context, they indicate the reciprocity, internal to structural causality, between independence and dependence, between imbrication and autonomy in the development of each instance. The point that is left unexplained here is the exact relationship between overand underdetermination, the question being whether one of the two can be considered primary. Could not underdetermination be the effect of the overdetermination itself? Or could we not think that overdetermination (which is by definition uneven) is possible because there is a constitutive underdetermination? The conclusion that we could draw from the very impossibility of answering these questions is perhaps more interesting: the key point here is that Althusser, by not establishing a hierarchy between them, is postulating both overand underdetermination as co-originary, as it were; thus, they are both included in the concept of structural causality in the same way, without there being any precedence of one over the other. 
The crucial consequence that Althusser draws from this conception of differential time is that the concept of ‘present’ must be radically reformulated. Althusser substitutes for the idea of the present as an ‘instant’ the concept of ‘conjuncture’, which becomes the name of the ‘present’ within the framework of structural causality. The present is not a simple moment, but is in itself a complex moment, ‘a time of times’, a time whose characteristic is to be ‘nonpresent’ to itself. As such, the idea of a complex and multiple present has an important consequence. On the one hand, it opposes the idea, which is linked by Althusser to the empirical conception of time (of which the model is Aristotle’s Physics), according to which time is a linear succession of instants that are in themselves simple; on the other, it also refuses the idea that each present of the ‘whole’ is a ‘full’ present, a present where all the elements coexist expressing one another.
It is this play of overand underdetermination that grounds the role of ‘absence’ in the scientific knowledge of the conjuncture. Althusser writes:
the present of one level is, so to speak, the absence of another, and this co-existence of a ‘presence’ and absences is simply the effect of the structure of the whole in its articulated decentricity. What is thus grasped as absences in a localized presence is precisely the non-localization of the structure of the whole. 
In this passage, the notion of absence is clearly admitted. There is, however, a crucial difference with respect to the absence that we find in ‘metonymic causality’. First, we must note that Althusser refers to a system of absences, rather than to the absence of a cause. Whereas for Miller the absence is the absence of the structure, Althusser is emphasizing the plurality of the absences, which correspond to the different and intertwined structures making up the complex of structural causality. Second, these ‘absences’ here are a function of the presence. It is not, in other words, an absence that is transcendent with respect to presence: ‘absence’ is the name of the ‘non-contemporaneity’ of the conjuncture – and not its principle of organization. Therefore, according to this passage, for Althusser any absence is always a function of presence, or, better said, the plurality of absences is the modality of the presence of the whole.
Determinate absence and the logic of irruption
How does Althusser’s insistence on presenceabsences link up with the problem of necessity and/ or contingency? As stated at the beginning, Reading Capital does not develop this point explicitly – we have only the symptomal identification après coup, in the introduction, of structural causality and ‘necessity of contingency’. It is in some letters to his psychoanalyst Diaktine, written in 1966, that Althusser broaches the problem of a logic capable of thinking the ‘birth’ of a novel structure, explicitly building on the concepts of non-contemporaneity and the system of absences as conceptualized in Reading Capital through structural causality.  Such logic is premissed upon a total refusal of the concept of genesis, which Althusser had already criticized in For Marx and Reading Capital, and the logic that such a concept grounds: one according to which the emergence of a certain structure, or of a certain phenomenon, depends on a linear development of a certain unity, which transforms itself into a new kind of unity. By contrast, the logic that Althusser seeks to extract from his concept of structural causality is one in which the transformation of the structure is not premissed upon a preliminary moment of totalization or unity, but rather upon the internal differentiation of the structure, or its non-homogeneity. He writes:
whereas the ideology of genesis presupposes that one can follow the trace of birth … and considers only what resembles the effect to be explained … this new logic can provoke the intervention of elements that may even seem to be absent from the conditions of a certain phenomenon. I believe that you will agree that absence possesses a certain efficacy, on the condition that it be not absence in general, or any other Heideggerian ‘openness’, but a determinate absence. 
The crucial term, here, is obviously ‘determinate absence’, and it suggests that Althusser is consciously rejecting the notion of an ‘absent cause’. It is from this conception that Althusser draws the consequences as to the irruption of a novel structure. By referring to Marx’s conceptualization of the transition from the feudal mode of production to capitalism, Althusser argues:
This structure cannot be thought, in its appearance [surgissement], as the effect of a filiation, but as the effect of a conjunction. This new logic has nothing to do with the linear causality of filiation, nor with Hegelian ‘dialectical’ logic, which only says out loud what is implicitly contained in the logic of linear causality. The elements defined by Marx [free labour-power, accumulated capital and technological inventions – SP] ‘combine’, or ‘conjoin’ by ‘taking hold’ in a new structure… What is important in Marx’s demonstration is that the three elements are not contemporary products of one and the same situation. 
What is crucial is that this theory of conjunction, based on the underdetermination and overdetermination of the whole (only the non-contemporaneity of the structure with itself allows for the production of ‘new elements’ that can eventually conjoin), is supposed to account for the ‘irruption’ of a new structure in such a way that the passage from one structure to another is not a necessary one. According to this logic, the elements that enter into the new combination that makes up the new structure are produced by the ‘old’ structure. But they are neither produced at the same time nor are products of the same genealogy (a term that Balibar had introduced in his contribution to Reading Capital);50 nor do they generate one another. Above all, they do not include the sufficient conditions of their conjunction, or of their ‘taking hold’, a concept that Althusser uses precisely to mean this conjunction/combination. In a sense, we might well say that, according to Althusser, feudalism posed the conditions of its own ‘overcoming’; but the conjunction, precisely, is what is not produced by feudalism itself as the outcome of its own internal laws. Thus, the production of a new ‘structure’ (its ‘taking hold’) is to be conceived as a real production: as Althusser suggested (in the passage from Reading Capital quoted earlier) we need to conceive ‘the historical relation between a result and its conditions of existence as a relation of production’, and therefore as ‘what we can call the necessity of its contingency’. In other words, these conditions become conditions of ‘overcoming’ only by virtue of a surplus of contingency. The same holds, evidently, for capitalism: to say that capitalism produces the elements that can lead to communism does not mean that capitalism is its own gravedigger: what is determinant is what capitalism does not ‘produce’ as its own results – that is, the combination of the elements in a new structure. From such a perspective, the passage from one ‘moment’ (in the Hegelian sense) to another is a contingent one. And it is in this sense that there is a necessity of contingency, a necessity of a surplus of contingency that brings together elements produced within the structure, in the internally underdetermined and overdetermined unity of the structure – this unity being a non-contemporaneous one.It is true that this aspect of the ‘surplus of contingency’ remains underdeveloped in these letters, in the sense that Althusser does not name it explicitly. We may therefore conclude that, just as Marx failed to produce explicitly the concepts that were nonetheless at work in Capital, Althusser failed to draw the conclusions implicit in the premisses of his argument, to which he, nonetheless, hinted at in the above-mentioned passage of Reading Capital. This is perhaps due to the fact that it is not that easy to find such a conceptualization of the ‘necessity of contingency’ in Marx, even if some passages – like the one commented upon by Althusser – may be thought to go in that direction. It is perhaps for this reason that Althusser, in the same months, thought that the thinker who provided the purest model of this ‘surplus of contingency’ was not Marx but Rousseau. In the notes from a course that Althusser was giving at the École normale supérieure in 1966 on the Second Discourse, he argues that in Rousseau it is possible to find a model of a causality without centre and without origin (a model in nuce of the concept of structural causality) as well as its counterpart: the idea that the ‘transition’ between an epoch and another in the unfolding of the historical process is premissed upon the irruption of a contingent accident. Commenting on Rousseau’s account of the unfolding of the historical process, Althusser notes:
the state of pure nature (I), the state of youth of the world (II), the state of war (III) do not hold in themselves any principle of resolution of their own contradiction … there had to occur some accidents to produce the transition from a state to another… For Rousseau every genesis is the transformation of a determinate contingency [d’une contingence] in necessity: that which comes about contingently [comme contingent] produces a new and irreversible necessity. Every necessity, conversely [inversement], has at its origin a determinate contingency. 
It is probably useless to point out that this reading of Rousseau, which, as always in Althusser, complements his reading of Marx,  pushes the concept of structural causality at the greatest possible distance from any idea of a logical or metaphysical necessity and stands in contrast to both Spinoza’s and Hegel’s necessity: there is neither a teleological necessity, nor a logical one. It is clear that Rousseau, among the authors he is in dialogue with at this time, provided Althusser with the purest example of a thought that promoted contingency – through the notions of ‘encounter’ and ‘accident’ – to the rank of what is determinant in the last instance in the transition from one ‘phase’, ‘structure’ or ‘mode of production’ to another. Contingency is not, as it was for Spinoza and for Hegel, an epistemological weakness of human knowledge; it is ontologically constitutive of history.
By 1966 Althusser had developed a conception of structural causality that, building on the ideas of non-contemporaneity and system of absences, had come to incorporate contingency as an essential dimension. The internal differentiation of the present in the overdetermination and underdetermination of the process, which is a process of processes, a structure of structures, may well pose the conditions of the replacement of a structure with another. Such a replacement does not rely on a moment of totalization, but is premissed upon a necessary surplus of contingency that can make some elements ‘conjoin’. Far from being an ‘implacable logic’ of the timeless reproduction of a given structure, as it has been read, structural causality was Althusser’s attempt to think through the category of ‘necessity of contingency’, and to provide Marxism with a category by means of which to think of history – its reproduction and also its ruptures – without resorting to the concept of contradiction. Althusser would criticize himself for his own detour through Spinoza a few years later, arguing that the price of that detour – the loss of the concept of contradiction – had been too high. However, what we find here (provided that we read structural causality without Lacan and without structuralism) is a concept that has not been fully explored in its potentialities
1. ^ L. Althusser and É. Balibar, Reading Capital , trans. Ben Brewster, Verso, London and New York, 2009, pp. 200–214 – henceforth RC. This is the translation of the second French edition of Lire le Capital, which included only Althusser’s and Balibar’s contributions in a revised version. In the course of this article I will refer to some passages of the first edition, available as L. Althusser et al., Lire le Capital , PUF, Paris, 1996 – henceforth LC. The English translation of the first edition, L. Althusser et al., Reading Capital: The Complete Edition , trans. Ben Brewster and David Fernbach, Verso, London and New York, 2016, had yet to be published when this article was written.
2. ^ RC, p. 17. 3. See A. Schmidt, History and Structure: An Essay on HegelianMarxist and Structuralist Theories of History , trans. Jeffrey Herf, MIT Press, Cambridge MA and London, 1981; E.P. Thompson, The Poverty of Theory and Other Essays, Merlin, London, 1978.
4. ^ Structural causality has been fiercely opposed by Hindess and Hirst in A. Cutler et al., Marx’s Capital and Capitalism Today, vol. 1, Routledge & Kegan Paul, London, 1977, p. 222ff. Laclau and Mouffe criticize Althusser’s Spinozist rationalism in Reading Capital in E. Laclau and C. Mouffe, Hegemony and Socialist Strategy: Towards a Radical Democratic Politics, Verso, London, 1985, pp. 98ff.
5. ^ Cf. P. Anderson, Considerations on Western Marxism, Verso, London, 1976, cited in P. Thomas, ‘Philosophical Strategies: Althusser and Spinoza’, Historical Materialism, vol. 10, no. 3, 2002, p. 80.
6. ^ See Thomas, ‘Philosophical Strategies’, p. 80. Cf. G. Elliott, Althusser: The Detour of Theory, Haymarket Books, Chicago, 2009. p. 150.
7. ^ See the writings col ected in L. Althusser, Philosophy of the Encounter, trans. G.M. Goshgarian, Verso, London and New York, 2006.
8. ^ L. Althusser, ‘On Content in the Thought of G.W. F. Hegel’ , in The Spectre of Hegel, trans. G.M. Goshgarian, Verso, London, 2014, pp. 17–172 – henceforth OC.
9. ^ Cf. the judgement of F. Matheron in the presentation of the first volume of L. Althusser, Écrits philosophiques et politiques, vol. I, ed. F. Matheron, STOCK/IMEC, Paris, 1994, pp. 13, 17 – henceforth EPP I. Matheron’s thesis is confirmed by G. Rametta, to whom we owe, to my knowledge, the most thorough reading of Althusser’s dissertation on Hegel. See G. Rametta, ‘Il contenuto è sempre giovane. La tesi su Hegel del giovane Althusser’, in G. Rametta, Metamorfosi del trascendentale. Percorsi multipli da Kant a Deleuze, Cleup, Padua, 2008, pp. 85–114.
10. ^ ‘The Hegelian concept is pure interiority … “The self has no outside” … The Hegelian concept is the movement through which the result recovers its origins by internalizing them; by revealing itself to be the origin of the origin. This process of envelopment implies that the initial term and the reflected term are aufgehoben in the result: that is, preserved and justified.’ OC, pp. 82–84, 94.
11. ^ ‘The concept pretends to externalize itself and posits differences which are, apart from its act of positing them, nothing at al , and which are therefore not real, but accidental … in order to grasp the insubstantiality of their substantiality, we need to look at things through the eyes of God, who sees the differences men seek desperately to experience disappear even before they have come about. To discover this reassuring [apaisante] perspective we have to situate ourselves, by means of philosophy, at the origin of the concept.’ OC, pp. 135–6.
12. ^ Ibid., pp. 142, 148.
13. ^ Ibid., p. 170. 14. Ibid., p. 168; my emphasis.
15. ^ Ibid., p. 167. 16. Ibid., p. 170. 17. See ‘Les problèmes de la philosophie de l’histoire (1955–1956)’, in L. Althusser, Politique et Histoire de Machiavel à Marx. Cours à l’École normale supérieur 1955–1972, ed. François Matheron, Éditions du Seuil, Paris, 2006, pp. 33–160 – henceforth PH.
18. ^ PH, p. 152. In this course, the ‘break’ between Hegel and the whole tradition of the philosophy of history and Marx hinges on the dichotomy closure/openness.
19. ^ Ibid., p. 179. 20. Ibid., p. 171. 21. L. Althusser, ‘Sur l’objectivité de l’histoire’ , in Solitude de Machiavel et autres textes, ed. Yves Sintomer, PUF, Paris, 1998, p. 28.
22. ^ Ibid., p. 26. 23. In the course on the philosophy of history Aron is cited several times. Althusser had read his Introduction à la philosophie de l’histoire (Gal imard, Paris, 1948) in 1953. According to F. Matheron and Y. Ichida, his notes on Aron’s book show that Althusser was particularly interested in the notion of hasard that Aron found in Cournot – a fact that confirms that Althusser was interested in developing a concept of causality capable of addressing the issue of contingency already in these years. See F. Matheron and Y. Ichida, ‘Althusser, un “typapart” une bibliothèque à part?’, Le Temps Modernes, vol. 664 no. 3, 2011, p. 206.
24. ^ L. Althusser, ‘Montesquieu: Politics and History’ , in Politics and History: Montesquieu, Rousseau, Marx, trans. Ben Brewster, Verso, London and New York, 2007, pp. 13–112 – henceforth MPH.
25. ^ MPH, p. 50. 26. Ibid., p. 45. 27. Ibid., p. 54. 28. ‘What is indeed remarkable in these factors, which either determine the very nature of the government (e.g. geographical extent, climate, soil) or a certain number of its laws, is the fact that they only act on their object indirectly’ (MPH, p. 55). In the 1956 course on the philosophy of history to which I referred above, Althusser argued that these factors determine directly the nature of the government, representing a mechanical causality (PH, p. 48). This shift in the reading of Montesquieu is crucial.
29. ^ G.M. Goshgarian drew the attention to Althusser’s use of the term ‘encounter’ in his book on Montesquieu in his ‘The Very Essence of the Object, the Soul of Marxism and Other Singular Things: Spinoza in Althusser 1959–67’, in K. Diefenbach et al. (ed.), Encountering Althusser, Bloomsbury, London and New York, 2013, pp. 89–111 (see pp. 90–94).
30. ^ MPH, p. 56, trans. modified.
31. ^ I take into account here only Althusser’s contributions.
32. ^ Cf. for instance RC, p. 133, where Althusser argues that ‘the project of a structural history poses serious problems’, as in structuralism (more accurately, in Lévi-Strauss) the destructuration and restructuration of the synchronic is down to ‘accidents’ that happen for purely ‘contingent reasons’ that can only be thought as pure ‘miracles’. Yet this passage should be understood more as a critique of the opposition between necessity (structure) and contingency, rather than as a critique of contingency per se. Althusser’s critique is here level ed against the structuralist idea that the synchronic is self-contemporaneous and internally undifferentiated; thus, the transition from a certain structure to another (which is the concern of structural history) could only be explained by appealing to a contingent event coming from outside the structure itself. As we shall see, Althusser’s theory of ‘necessity of contingency’ wil rely on the internal ‘non-coincidence’ of the ‘structure’ with itself.
33. ^ RC, pp. 47–8; my emphasis.
34. ^ ‘The rationality of the philosophy of Enlightenment to which Hegel gave the systematic form of the development of the concept is merely an ideological conception both of reason and of its history’ (RC, p. 47). Althusser examines this continuity at length in ‘Les problèmes de la philosophie de l’histoire (1955–1956)’, pp. 33–160.
35. ^ Spinoza, Ethics, I, Proposition XXIX.
36. ^ It was Étienne Balibar who was the first to draw attention to the notion of ‘underdetermination’ in Althusser. See his ‘Avant-propos pour la réédition de 1996’, in L. Althusser, Pour Marx , La Découverte, Paris, 2005, p. xi i.
37. ^ ‘Let us assume the presence of an element that turns back on reality and perceives it, reflects it and signifies it, an element capable of redoubling itself on its own account … From the moment that the structure involves the element we have mentioned: – its actuality becomes an experience or experiment; – the virtuality of the structuring [le structurant] is converted into an absence; – this absence is produced in the real order of the structure: the action of the structure comes to be supported by a lack. The structuring [le structurant], by not being there, governs the real.’ J.-A. Mil er, ‘Action de la structure’, Cahiers pour l’Analyse 9, 1968, pp. 93–105; English translation: http:// cahiers.kingston.ac.uk/pdf/cpa9.6.mil er.translation.pdf.
38. ^ ‘By means of what concept is it possible to think the new type of determination which has just been identified as the determination of the phenomena of a given region by the structure of that region? More general y, by means of what concept, or what set of concepts, is it possible to think the determination of the elements of a structure, and the structural relations between those elements, and al the effects of those relations by the effectivity of that structure? And, a fortiori, by means of what concept or what set of concepts to think the determination of a subordinate structure by a dominant structure; in other words, how is it possible to define the concept of a structural causality?’ RC, pp. 205–6.
39. ^ LC, p. 646, variant
62. ^ In RC, p. 209, Althusser insists that ‘the absence of the cause in the structure’s metonymic causality on its effects … is the very form of the interiority of the structure … it implies that the structure is immanent in its effects in the Spinozist sense of the term.’
40. ^ On this, see also the analyses developed by Y. Sato in his remarkable Pouvoir et résistance, L’Harmattan, Paris, 2007, pp. 173 ff., to which I am indebted. For a different reading, see W. Montag, Althusser and His Contemporaries, Duke University Press, Durham NC and London, pp. 81–100.
41. ^ RC, p. 101–31.
42. ^ RC, pp. 110–11. It is important to note that in this chapter the notion of a determination in the last instance by economy does not play (contrary to what Laclau and Mouffe suggested) any conceptual role. Althusser always refers to the totality of the levels. 43. RC, p. 118. 44. Cf. L. Althusser, Essays in Self-Criticism, trans. Graham Lock, NLB, London, 1976, p. 187.
45. ^ If we compare this reciprocity and its consequences with the way in which overdetermination was introduced in For Marx (and especial y in ‘Contradiction and Overdetermination’ and ‘On the Materialist Dialectic’), it seems that, by introducing the underdetermination, Althusser is stressing that the relations between the levels cannot be reduced to the logic of the conditions of existence. Effectively, in those two essays the overdetermination of the economic contradiction was due to the superstructures being also the conditions of existence of economy itself. But in ‘Contradiction and Overdetermination’ Althusser also said that for the main contradiction to be activated an intervention of ‘currents’ foreign to the task of revolution was needed (L. Althusser, For Marx , trans. Ben Brewster, Verso, London, 2005, p. 99). Now, on the basis of overdetermination, how are these ‘paradoxical y foreign’ currents understandable? It may be said that the concept of underdetermination accounts for the existence, in the social whole, of elements and currents that are not reducible to the status of conditions of existence; the various levels or instances (thus, the various contradictions), are certainly for Althusser also the conditions of existence of the economic level, but they are not only that. Each and every level has its own history and time, which is subject to a torsion by its articulation upon other levels, but is also relatively independent; the degree of its torsion is not forever determined or determinable in advance.
46. ^ RC, pp. 115–16.
47. ^ L. Althusser, ‘Letters to D.’, in Writings on Psychoanalysis , ed. Oliver Corpet and François Matheron, trans. Jeffrey Mehlman, Columbia University Press, New York, 1996, pp. 33–78 – hereafter LD.
48. ^ LD, p. 61. 49. L. Althusser, ‘On Genesis’, Décalages, vol. 1, no. 2, 2013.
Available at: http://scholar.oxy.edu/decalages/vol1/iss2/11.Thisbrief [archive] text was meant to be another letter to Diaktine.
50. ^ RC, p. 317. 51. PH, p. 308. 52. Of course, in Rousseau, Althusser does not find the concepts of overand under determination. What I am suggesting is that he finds in Rousseau a model that stresses the moment of the ‘contingent taking hold’.