Translation of F. Laruelle’s “The Transcendental Computer: A Non-Philosophical Utopia”
Source: https://speculativeheresy.wordpress.com Translation of F. Laruelle’s “The Transcendental Computer: A Non-Philosophical Utopia”
The following is a revision and correction of a preliminary draft drawn up by my friend Chris Eby several years ago. Since he undertook the task for the sheer pleasure of it, it’s state on Scribd cried out for a new edition. I have taken the liberty of reworking the essay in its entirety, so now all problems with the translation rest on my shoulders!
The Transcendental Computer
A Non-Philosophical Utopia
Translated by Taylor Adkins and Chris Eby
The unified theory of thought and computing [calcul], a unification in-the-last-identity, is a task facing every encyclopedic mind (Morin, Serres). It is also the theme of the transcendental computer (TC), of a machine that would have a transcendental relation to philosophy in its entirety and would therefore be able to compute-think the blendings of thought and computing according to a “unified” mode, such as, for example, a transcendental arithmetic like Platonism or any other combination of these prevalent terms in philosophy and computing. Beforehand, a prejudicial question concerning the degree of non-philosophy’s automaticity should be dealt with. In this sense, what follows is an attempt at the limits of the theme of a transcendental computer.
Automaton and Unimaton
If non-philosophical practice is localized in or measured by an effect (unification-effect or clone-effect), then, since the One-in-person cannot be this reference point due to being unidentifiable, this effect is a type of possible disalienation for statements of the World, which are de jure representable or philosophizable. It could already be objected, and one could already understand from what will be said, that this non-philosophical practice is not identifiable as specifically “non-philosophical,” since its cause is Identity-in-person, which is not identifiable in exteriority as an available criterion because it is lived-in-immanence. Its cause after all could be just as much the effect of a machine simulating a completely absent subject, a sort of charitable automatism (automatic operation), and then, all things considered, it would be unnecessary to critique philosophy. If a machine in the classical sense of this word can do what non-philosophy does, aren’t non-philosophy and its theory useless? More exactly, the objection is the consequence or the result of a division of non-philosophy liable to give rise to two images: an inert theoretical image of a machine or mechanism composed of objectified parts, and an image of practical functionality, the latter image given without distance of objectification, but lived. The objection supposes the right to resolve non-philosophical practice into an inert structure (of photographing), which supposes that this structure must be constructed beforehand, even before it is able to function and for it to be able to function. However, if there is indeed a presupposed to this thought, it’s not a structure, a schema, a legible figure in a space of transcendence (this type of structure exists for philosophy itself, it’s one of philosophy’s modes of givenness, which is itself philosophical), it’s precisely what defies every transcendence and every inert structure composed of terms and relations, points and vectors, etc.; it’s the Real-in-person.
According to this objection, the effects alone would be appreciable as arising or not from non-philosophy’s presupposeds of “humanhood” [humanéité]. But this objection conceals its own presupposition, which is to consider a priori effects themselves as already being those of an automatism (and to consider this automatism and its effects as insertable into a structure), i.e. to consider inert effects registered and given in the World (Husserl would have said in the “natural attitude”). This is to prejudge the nature of Identity-in-person and, based on this vicious circle, to conclude from the automatic nature and supposed mechanism of these effects to their producibility by a machine. But nothing happens like this in non-philosophical practice, which is not an automaton simulating man. That would instead be a contemplative or theoreticist vision, and thus an objectivism that would be absolute, materialist or strictly mechanistic in an a priori manner, which is a philosophical possibility. Let’s examine these effects of non-philosophical practice, since this is what’s behind the objection’s secret argument or criterion.
Of course there are effects (of cloning), but there’s nothing automatic about them, and they shouldn’t be considered dogmatically as inert things. They are partly, on the material side, formed by the World’s inert things but already enveloped in a horizon of philosophizability or transcendence (which itself includes the possibility of a subject or whatever goes by that name), since these two components are blended. And another “part” of them is formed by a radical or “lived” immanence (the Lived-in-person) that excludes being blended and is not in any way a part of a whole. Yet the blending and non-blending do not blend together but, so to speak, wed or embrace one another, join together without a synthetic band in an irreversible alliance known as the existing-Stranger-subject or subject-existing-in-struggle. This subject is the veritable effect, entire in its identity and its (unilateral) duality: this is practice as including a material of philosophy-form. Where is the effect of practice legible in the subject? In and as what we have called the phenomenon or the appearing of the practical subject, which is not the juxtaposition of two halves but the transformation of one of the sides by the other with which it is unified without synthesis. If, for example, the effects of textual statements are not continuously related to Identity-in-person or cloned, performed, then they again give rise to practice as thingified and inert, automatic, i.e. perceived from philosophy’s point of view alone and then left to the division that we mentioned above. The being-performed of Identity or of Man-in-person and of the latter’s cloned effects, however, is not itself visible or sensible, but leaves its mark through such effects not just in the visible and the sensible of history and the world, as in a receptacle, but directly on their philosophy- or world-form. Transformed in this way, this form gives rise to an appearing that is phenomenal or determined-in-the-last-identity (in-the-last-lived-experience?). This is philosophy-form as such given in-One, or better still, its transcendental identity.
Subjectivity, whose phenomenal appearing in-One is the fabric of the subject effect, is included in this philosophy-form in particular. This fully excludes that the determined effects, which are of philosophical extraction, can be produced by an automatic system, at least provided that philosophy’s transcendental mechanism can itself escape from this automatism and this reduction to a simple mechanism. It’s transcendence in general that excludes philosophy’s reduction to an algorithm. Now one can obviously pose the problem of the possible degree of automation of transcendence, which is philosophy’s transcendental nerve. But to the extent that it continues on, albeit transformed, in the subject, philosophy limits the chances of automaticity and formalism.
One can obviously compare the modes of immanence of Man and machine. Or instead the latter supposes a human whose functionality the machine imitates to the nearest degree; this is an interiority of consciousness spread out in space. Or instead this machinic and algorithmic immanence is first, and it’s consciousness or our concept of consciousness that imitates the machine. One turns in a vicious circle.
Man-in-person indeed is not a subject in the traditional sense or a “man” in the anthropological sense, a mode of consciousness or being in general. In a sense the “passivity” of Man-in-person does nothing but reinforce the “mechanistic” aspect, even if it be said that it’s of pure lived experience. Its aspect of automatism is perhaps an appearance created by the absence or the lack of an active, localizable and identifiable subject, which makes one believe in a machine. Identity-in-person resembles a machine without being a machine; here what makes one think of transcendence and of its void of subjectivity is radical immanence. Radical immanence is also devoid of subjectivity but not of lived experience: that’s what distinguishes radical immanence from a machine. Here it’s not the machine that simulates a man at the vanishing limit of consciousness, but Man-in-person that simulates a machine or an automatism.
Man, being neither a consciousness nor an unconscious, undoubtedly and in a negative way seems closer to the machine, if not its immanence, and is necessary qua presupposed, a logical and real necessity without blending. Everything that comes from philosophy or supposes it is of the order of the Real, at least as symptom; what comes from logic and necessity is of the order of identity. One could say that Man-in-person is an-axiomatic or an-hypothetical, in the sense that the privative “an-” is radical or expresses that Man is in-Man and not to (him) self or in (him) self, and thus is foreclosed to the philosopher and to all automaticity. Instead of supposing axioms true as in logic, one supposes them real or anaxiomatic. Not the axiom’s axiom, but a non-axiom or an-axiomatic axiom.
These are unilateral axioms, and they are unilateral only due to one of their sides; these are therefore not auto-referential axioms (non-Gödelism), although it’s not certain that such a thing exists, save in language and metalanguage form, since metadiscourse serves to articulate the axioms or their status. The One-in-One is not the 1 facing philosophy’s 2/3. It’s not describable in terms of absolute transcendence but by axioms that give where its effects are. Even qua automatism, the One is perceptible only through these effects of discourse or its practice, not in itself, since it’s not a thing or intellectual intuition. Michel Henry couldn’t keep himself from giving it an identifiable content in transcendence. But this isn’t algorithmic automatism, which is integrally visible and given in a finitary and quasi-geometrical way. Scientific automatism is that of transcendence, but it’s not philosophical; it thus supposes a metalanguage and is undoubtedly the complex form of the scientific relation to the real.
Man-in-person is neither an auto-maton, an auto-nomic auto-functional functionality, nor a functionality that supposes a multiplicity of pieces and effects. In all due rigor, Man-in-person is precisely a uni-maton determining a practice (uni-maton signifies that the “-maton” aspect is aligned with the “identity” aspect or determined-in-One). The term “immanence” is ultimately misleading, like all the others, and makes philosophers believe in a thing, whereas here it is like all the rest nothing but an attribute that disappears into an axiom which uses it, a term that designates the Real by objective appearance. Non-philosophical practice is indeed a uni-maton in the sense in which the latter is a unified first term and not a unitary syn-tagm. However, it can be nothing but the condition of knowledge for a philosophizable kind of automatism at best. The latter claims to be in “auto-” mode (which is never completely true). The auto- supposes an active-passive immanence, a transcendence, a unified system of multiple pieces, at least two pieces, and ultimately 2/3.
Ultimately one must begin by distinguishing between two forms of automatism, the philosophical and the logical, and a minimal form that is instead unimatic. The logical admits a metalanguage, the philosophical, instead, a hermeneutics; the unimatic prohibits metalanguage and hermeneutics or carries out their unified theory. In all three cases it is a matter of speaking “about” a discipline, philosophy, logic. These latter two resolve the problem by speaking of one another with their own language, which also allows them, obviously, to speak of themselves respectively.
Since the duality of logical metalanguage is opposed to philosophical blending, non-philosophy is perhaps what unifies these two practices, the transcendental and the metalinguistic, two types of duality, or better yet what I have always called the philosophical posture and the scientific posture. Since these would be the three grand styles, perhaps the word style is the best? What I call axiom is neither a metalanguage for philosophy and its own “axioms” nor a philosophical hermeneutics wherein something transcendental is conserved, even if the axioms take from metalanguage and the interpretation of philosophical postulates. The anaxiomatic or non-axiomatic Real prevents axioms from sinking into Being, Nothingness, the Multiple, ontology, or into logic’s finitary-intuitive space and prevents them from symbolizing with ideality. It withdraws them from their sufficiency.
The last twist on the problem who simulates whom? requires seeing that this simulation exists for a philosopher, not for human Identity itself which knows itself separated, from the machine as well as the rest. It would obviously be necessary to review the concept of simulation in all its usages and perhaps reverse the sense of this obsession with the machine. Ultimately isn’t it the philosophized machine or machine in-philosophy, which is what’s being talked about, that provides the fantasy of a simulation of the machine through Identity-in-person? Would there be a narcissism of the philosophized machine that would throw the simulation operation back onto Man-in-person? Watch as I am beautiful and fascinating…
Philosophy-form has not become a pure machine once it’s reduced to the state of symptom. In this case it’s Man-in-person who would reduce this form to the state of automatism, while in addition Man-in-person and its practice would arise from the uni-maton. Non-philosophy or the unimaton would utilize philosophy by reducing it as automaton in a truly special sense, but whence would the impression come that non-philosophy does things that a machine could do?
The thesis of the possibility of a transcendental computer (TC) could be sustained in two distinct forms:
— in a strictly machinic and technological type of AI form (artificial intelligence), immediately supposed realizable in the near technological future, with no other difficulty than that of the current age,
— in a non-philosophical form for which a TC is a similar but indirect Idea that supposes a detour out of the machine. This bridge between the machine and the transcendental is the unified-in-the-last-instance theory of thought and computing. This goes without saying that the machine’s conditions are necessary but insufficient, and thus that a machine alone cannot be a TC but that Man is necessary for this (not as consciousness, which eliminates a part of the classical discussions between philosophers and computer scientists of AI, for we no longer fully oppose thought to computing).
Solution 1 would realize the same performances as the TC of solution 2. This implies that it would suppose that its machine attains the same effects as Real structure + Determination-in-the-last-instance (DLI). Can a machine imitate immanence and above all DLI? It’s at least doubtful. If one refuses to make this postulation (since the type of effects produced by DLI necessarily suppose this immanence, which cannot be simulated simply to the point of being mistaken), one is, however, obliged to suppose or to be given one third of synthesis between philosophy and the machine, which is the concept of performance (“the same performances”).
It is necessary to orient the discussion around the catchall concept of performance, which in general allows for AI to claim to equal the “performances” of intelligence and even thought (for the moment we are not yet distinguishing between the two). Performance is measurable and utilized or supposed as one of computing’s criteria of identification with thought, and inversely as the criterion of thought’s reduction to computing. However, the psychological situation is even more complex, since the machine does not attain the same performances as “human” intelligence except on condition of surpassing them or secretly hoping to surpass them more or less. If not, what’s the benefit or the point? (unless to suppose that it’s the intelligence itself which always wants to surpass itself by creating the machine’s mirror in which it can witness itself triumphing over itself)?
The notion of performance is a presupposed that encroaches on the meaning of intelligence and of what it can do. This is a notion of technological and quantitative measure but which is supposed valid for intelligence. It supposes between the departure and the target an identity of effects or ends and undoubtedly a homogeneity of syntax and semantics, an algorithmic transparency. This is to say that it’s worth nothing in philosophy (as much a failure as a success, and failure here is not necessarily the opposite of success) for which such a transparency does not exist, since philosophy determines dualities reciprocally, for example its syntaxes and its matters. Here, one must distinguish between intelligence and philosophy. “Cognition” is a priori parceled into more or less closed and isolated systems that in effect can be measured in terms of performance. AI prejudges intelligence, what intelligence can do by setting for it limits or goals (determined and finite in the measurable sense) in order to compare it to the machine. With philosophy, everything runs in another manner. One could even define intelligence rigorously by the type of performance that a machine can simulate, either in its functionality or in its effects. But philosophy cannot be reduced in this way a priori, i.e. parceled into functions or effects and prejudged. Why? Philosophy uses intelligence or cognition but on behalf of a special form of thought, probably irreducible to any numerical combination. No doubt many objects or operations “of” philosophy are therefore reducible to performances, but they are in reality intra-philosophical and hearken back to an operative horizon that is forgotten on principle and strictly cannot be “recalled” by computing. This transcendental horizon is auto-position or “philosophical decision.” Auto-position seems like a goal to be attained, and one that philosophy attains, but philosophy attains auto-position as much as it misses it or at least includes its misfire in its success. Auto-position is a superior performance or the “superior” and transcendental concept of performance. The schema in 2/3 or 3/2 is an arithmetical approximation, whereas philosophy is a transcendental arithmetic or is valid for existence or the real. Arithmetic also “equals” the real, but a region of the real and not fundamentally the real itself, and moreover is valid for it or possesses a constituting power of legislation. Philosophy is transcendental in a narrow sense for experience, and in a broader sense for itself, insofar as it is sometimes thought of the real but also the real or thought as real. Yet this relation to experience and/or to itself is called transcendental because it conditions or legislates on its object, to which it appears at the same time to be neither exhausted nor reduced thereby. Here the concept of performance therefore only has a local and not global, provisional but not final, sense. Wouldn’t this paradoxically be an artifact or a concept, a representation of consciousness?
How does one imagine that the act of position, which has both a status of metaphor and of proper sense (here the proper or the real must exist in philosophy and it must not be metaphorical through and through, even if it is revealed to be hallucinatory under other conditions), can be computable, reducible to effects of numerical combinations? Even more so, how does one imagine that the division and duplication of position, the acts of de-position and over-position, in short the “auto,” can be computable? One last argument of the same type can be grounded on philosophy’s auto-speculative kernel qua speculation. Philosophical specularity (the foundation of its theoreticism) is not simple; there must be a mirror for it that can take the place of the real and in certain “idealist” cases can itself be grasped in the play of reflections. This ultimate structure of philosophy, presupposed by the doctrines that claim philosophy as a reference but do not pursue its analysis up to its last or minimal end, is a phenomenon that could be called qualitative, at least as much as certain thinkers could want it to be quantitative or simply to derive it as inessential. Philosophy’s grand law, the law that is such that philosophy submits to it, is to be a blending of the numerical and the qualitative, here in the form of position or specularity. Nothing authorizes a philosopher, i.e. someone who distinguishes between philosophy and cognition, to let himself be intimidated by the machine’s performances, which are truly performances but nothing more.
The philosopher must seem to allow the machine to grow and even to make it grow where he can, accepting, more or less approximately, to empty philosophy of its substance of intelligence. But a residue survives this cognitivist reduction, which is the first and last, numerically invulnerable kernel. Why does one want to save this envelope that philosophers themselves ostensibly create to forget? It merits being saved if it is as original and specific, incalculable, as it is probable. Even Badiou, who develops an ontology of the “pure” numerical, reserves philosophy’s role as the power of gathering, and thus of quasi synthesis or system, a sort of complement or supplement to mathematics. Furthermore, the duality of the numerical and the continuous, of the mathematical and the philosophical (these terms should merit being nuanced and utilized with caution…), is a historical constant that traverses every occidental thought, wherein the numerical regularly announces its victory and the continuous its survival. In their generality, these are imaginary “transcendentals” or apparently inseparable paradigms (Bachelard), as if thought were condemned to follow a double path or to struggle on two fronts. These are good reasons for maintaining philosophy’s originality, at least that of its essence. Non-philosophy is among other things a manner of registering this survival without claiming to see one of the parts crush the other but by relating each to an instance that is neither the continuous (dominant in philosophy) nor the discontinuous (dominant in science).
The contradictory argumentation between AI and the advocates of Consciousness is always wearisome and the same. The first states that it has already realized such a performance and therefore that it will realize others still more important in the field of thought. It is animated by a philosophical claim but advances under the guise of science. The second always responds by one last domain in which it takes refuge in its mastery and challenges AI to access it. But this is always one of philosophy’s objects or domains, not philosophy in its essence. I will consider that both this conquest and this self-defense have a positivity and a validity, that they even only have sense due to their reciprocal opposition, and that this indeed testifies to their common claim, their will to the absolute, which this opposition divides. I propose to call this conflict the antithetic of cognition or of computing-thought, a restrained antithetic in Consciousness/Cognition form that is generalized or expanded in Philosophy (rather than thought)/Computing form. It will be posited that non-philosophy is an attempt to give a (non-Kantian…) solution to this conflict, i.e. in order to “exit” it or more precisely to show how and under what conditions thought can never have entered it.
As for the program/execution distinction (and, on this model, theory/practice), it’s a duality of another nature, internal to computer engineering. In a sense every duality of this kind is always usable to characterize non-philosophy, which functions with such dualities but on condition of previously interpreting each duality in a philosophical rather than unilaterally machinic sense, i.e. on condition of deploying their potential horizon of philosophical sense. Non-philosophy only denies “over-human” or “ultra-human” claims, but it is a pragmatics that can make good use of all dualities. If this preliminary preparation phase of the material is not gone through, one reduces philosophy and non-philosophy inversely to inert sets and one forgets what truly constitutes “life,” which is perhaps hallucinatory but philosophy’s life all the same, including auto-position, not to mention non-philosophy’s life, vision-in-One. One can believe to have resolved the TC in a purely machinic manner if one begins by reducing or restricting the problem’s extension and its givens in the program/execution couple. Transcendental life and, even less so, real lived experience are not reducible to algorithmic repetitions but can make use of them (always unilaterality…).
A performance fundamentally consists in simulating either a functionality or more simply the effects (“the same effects,” but one last simulation is concealed in this notion), by doing as well as…succeeding in an already defined or fixed task, however much it surpasses the latter. But who has completed the task or determined the goal to be attained, and therefore who has already realized it in a certain way? This question does not have meaning for numerical representation but has a fundamental meaning for philosophy, which realizes or effectuates things for the first time, which is first philosophy or radical commencement. Even if that would be a claim, that’s what the meaning of philosophy and of its life is, or of its “functionality”: it’s undoubtedly a repetition, but second or in relation to itself, an auto-repetition, and thus ultimately first. Philosophy is first, engineered computing or the machinic use of computing (I’m not speaking of arithmetic but of its usage in AI, a “usage” which should already attract attention to the degree that there is philosophizable virtuality in this notion) imitates or simulates something other than itself. Philosophy is not a performance, neither a simple machine despite “desiring machines,” nor even a “comporting” despite the Verhalten of Heideggerian Dasein, which are intra-philosophical interpretations impregnated with metaphor and thus inseparable from language.
If philosophy is not reduced to Consciousness and to its… “performances” and is revealed to be all the more irreducible to a machine using computing, non-philosophy radicalizes this irreducibility. Just as the Lived-without-life radicalizes Life (one of philosophy’s regular transcendental themes), the Performed-without-performation (and hence, more so, without-performance) radicalizes the concepts of performativity and performance. The Performed-without-performation is the first name or symbol (already an axiom) that grounds the critique of sufficiency, which quite visibly impregnates the notion of performance, but without simply denying it or conflicting with it.
What Simulates What, Non-philosophy or the Machine?
It is undoubtedly this refusal of the Consciousness/AI antithetic that gives the impression that non-philosophy is better prepared than philosophy to knot together “amicable” (Heidegger) relations with computing and more generally every form of automaticity. Non-philosophy can appear like an attempt to save philosophy against or “from” its traditional adversaries, but that would be merely a consequence, and the attempt at the solution of the antithetic merely an effect, not a cause or a motif of non-philosophy. The resistance in philosophy that non-philosophy critiques fully exceeds philosophy’s resistance to cognitivism. But the most expanded concept of philosophy is required in order to reveal the force and resistance, perhaps the source, of the continuous or the analogical. Let’s attempt to uncover the reason for this broader proximity and what obliges us to avoid believing in a seemingly possible computerized reduction of non-philosophy.
The Performed is not defined by the doing-saying couple in the style of linguistic performativity, but as that which determines in-the-last-identity the blending of performation and the performed. This type of Real seems at first sight to require us to rid ourselves of philosophy, even though it only rids us of Consciousness at best, and therefore requires us to be able to simulate the machine or simply the Unconscious. It is not said so easily that philosophy simulates the machine, but one is more easily tempted to say this of non-philosophy. This is because the Performed or Man-in-person seems to be a standstill, an ontological or even formal void or a blank screen. Hence the impression that non-philosophy is an automatism and above all a machine. But nothingness or even the void can be defined ontologically, not the Performed. Non-consistency is what’s important; it’s not more Nothingness than Being but determines their blending; it’s non-nothingness, the (non-) One such that it applies just as much to non-being, i.e. to nothingness. Just because it’s a “negative condition” or sine qua non doesn’t make it a positive essence (= that without which); it’s a non-essence, a non-(that without which), which therefore determines but as a negative condition, necessary but without contributing any positive predicate to the material and to its positivity. The cause is positively or philosophically absent, but retracting it from this positivity doesn’t relegate it to nothingness. It is absent qua activity and passivity insofar as these are blended. Can one speak of a negative acting? No more than of a positive acting. Even the positive and the negative couple are not satisfactory if one claims to use them predicatively and with apophantic definitions. There is thus nothing positive in general about this “negative” trait, but it’s positive, so to speak, in its kind. It can therefore be said of the real cause that it either acts or does not act (neither is their synthesis or their “at the same time,” cf. Derrida)—this is its non-consistency, and the real cause completely determines the blending of acting and non-acting. “To determine,” in philosophy and under any positive material condition, is to affirm or to imprint real identity “negatively.” It appears to me that this manner of thinking, which undoubtedly can seem by its apparent dogmatism to bring non-philosophy and a certain scientific argumentation closer, is foreign both to philosophy and to science.
This effect is extended explicitly in the Stranger-subject. The clone, i.e. the transcendental phenomenon, is structured as One (of) philosophy, or as uni-lateral Identity. This structure at the outset makes the clone foreign to philosophy in itself, which is constructed at least on two basic sides. The One itself has no side, contrary to what M. Henry, who turns the One into a transcendental Ego, posits; the clone-Identity has a single side, philosophy in itself has 2/3 sides or thinks itself as 2/3. The trait of strangerhood [étrangèreté] no longer has anything to do with an otherness or a transcendence simply opposed to philosophy. There is transcendence of the two sides, which is necessary for there to be a certain efficacy or for the clone to cut out and into the World’s transcendence. But the two transcendences (which obviously contain correlative immanence) are heterogeneous structures, the philosophical in itself is bi-facial, the cloned is uni-facial. A machine is always bi-facial in each of its “pieces” and effects, i.e. multi-facial. The machine tends toward autonomy and wants to think like philosophy does by making a success of its tour de force; it pushes autonomy as far away as possible and stumbles on the machine’s agent manufacturer, but gets nearer to non-philosophy in so far as it has a presupposed. The idealistic argument according to which machines can build other machines does not, despite appearances, forget that a first constructor, an anthropomorphic inventor of the first machine, is necessary, but it can always hope to reduce this inventor in turn to a component inseparable from a continuous “man-machine system,” obviously at the risk of inciting protests from the rival party of Consciousness. On the other hand, it “forgets” something else, which is that man-machine systems tend toward the auto-dissolution of all their internal distinctions and toward inherent nihilism, and that if this phenomenon is only tendential, this is because there is an instance capable of re-determining them and re-launching them, so to speak. It is necessary to distinguish between an absolute commencement (thus relative-absolute) of the man-machine circuit that disappears in the system. And a radical commencement, a first techno-logy or a non-technology, a human subject in-the-last-identity but existing in accordance with variables that are technical discoveries: therefore, a human subjectivity but co-determined by the forms and the style of various technologies. This argument is apparently too simple and formal, but here there is also an antithetic of technology between those who want a first anthropological commencement of the tool circuit, a human agent, and those who, like Leibniz, infinitely prolong the circuit up to a God-machine or a universe-machine. Non-philosophy resolves this antithetic between the constructor man of consciousness and the machine of machines, by suggesting that its sense is purely apparent, indeed hallucinatory, and by relating it unilaterally to Man-without-machine, who determines a machine-thought qua clone of the techno-logical blending. This is to say that the hypotheses on the machine’s exact origin and power remain those of the metaphysical order, and thus their solution is not within our scope.
Don’t conflate the program (non-philosophy supposed achieved or in a stable state) with non-philosophy’s material. What one puts in the program is variable, provided that it has philosophy’s variance-and-invariance. The rules and procedure of unilateral duality are fixed once the material itself is given and fixed, since it intervenes in the formulation of the rules (which always have a concrete aspect). Under this condition of the material’s fixedness, non-philosophy is indeed a machine or regularly transforms a given material into a given product, and as a result can appear like a program that simply awaits its execution. It’s even a human machine, or a machine lived and determined in-the-last-instance by Man. Yet there is then something bizarre, close to science-fiction in this concept, as if a machine in good and due form, selected from a technological circuit, has been transplanted not into a Consciousness but into Man-in-person. Non-philosophy is no longer this monster obtained by synthesis of technology and the Real. Not to mention that the material’s fixation, once and for all, is a return to a philosophical gesture that equally fixes in turn and thus makes the Real transcendent. All is lost, but this would be a joke of science-fiction, in some sense a “radical” joke.
The material only varies, and with it the rules of unilateral duality in their formulation, if a transcendental indifference and equivalence, which suppose a radical immanent Real, of the materials are posited. When transcendence is the unique principle, the material’s contingency disappears and the process becomes fixed in a new circle, in the philosophical thesis or doctrine. It is necessary to oppose philosophy’s “once and for all” (cf. Deleuze) with non-philosophy’s “one time each time” and its special “performativity.” It’s lived experience or the Real in its radical identity that one time each time determines (without creating) the material (and its invariant form) and clones a subject from it. Thus the most “singular” identity is now said of totality or of wholes, therefore also of invariant phenomena (since these exist), which makes them foreign to philosophical and technological economy. Non-philosophy is a machine necessarily specified or even “singularized” (identified) qua machine by what information “enters” there, which is a necessity that in fact stems from its “negative” cause. Unilateral duality is indeed an invariant structure, but one must distinguish in this formulation between the invariance effect that comes from the surreptitious or senseless fixation of a philosophical vocabulary with its horizon of potentiality (an artifact-invariance), and a deeper invariance that is reduced in-the-last-instance to the cause’s identity-in-identity. As if (this is an objective effect or an objective appearance) philosophy’s invariance vanished here, became elusive and were no longer even identifiable and recognizable except by the invariance of philosophy-form and its content of terms or its “semantics.”
It is difficult under these conditions to make a program of non-philosophy in the computerized sense. Or in that case it is a program one time each time, the Program’s transcendental identity or clone. The whole chain of causes and effects (Real + DLI) is contaminated by the transcendental contingency (which comes from the Real) that affects philosophy’s variant-invariant form (with, in addition, the contingency of the last philosophizable, empirical thing). The formulations given of non-philosophy until now, for example here at present, if they are objective through and through at a given moment, can give the impression that it is a matter of a program to be executed. This is a theoreticist normalization of non-philosophy by the philosophical posture. This program’s objective appearance is not its essence, only its reification or its worldification in a T1 moment. If the given worldly or historical time is posited as the essentially determining affair, then philosophy returns through its intermediary. This is a contemplation of practice, with the latter always one time each time in its transcendental identity, but its contemplation denies or negates the character of the material’s radical-transcendental contingency. Syntax and material are already inseparable in philosophy (this is the transcendental as philosophy’s trait), and if this connection seems to slacken in non-philosophy, this is perhaps an illusion, because the independent cause of every material that it renders contingent turns this contingency into an imposed or forced negative necessity. One cannot separate or isolate pure, formal and algorithmically manipulable rules; non-philosophy solely has an algorithmic aspect (a transformed material) of the machine, even of the automaton, and it’s a machine indeed, but determined in-the-last-instance by Man.
François Laruelle, “L’ordinateur transcendantale: une utopie non-philosophique,” in Homo ex machina, ed. F. Laruelle (Paris: l’Harmattan, 2005). [TN]
 This word is quite broad, and means calculus and calculation, but also counting or arithmetic (informally), and computation. Here, although counting is implied and should be kept in mind, the notion of “computing” comes closest in theme to the title of the essay, i.e. the transcendental computer. However, it should be noted that the word for “computer” in French is “ordinateur”, which again points to the numerical aspects of counting (ordinal numbers). Thus to reflect the coordination of the terms “calcul” and ”ordinateur”, I have chosen to translate calcul in the loose, conventional and ordinary sense of “computing”, taken broadly. This is also mainly because calculating/counting does not get at the active, programmatic (literally) sense of the performances that computing and computers imply. [TN]
 French anneau de synthèse, the anneau or ring signifying a wedding ring. [TN]
 French soi, reflexive pronoun. [TN]
 French fantasmer. [TN]
 Strange phrase…It should be noted that the adjectives are feminine in gender here…So the I of the enunciation could either be taken as sexualized (a female voice), or the I could be that of radical immanence…(a feminine noun). But, perhaps more likely, since the idea concerns machine/man simulating one another (What simulates what?), the feminine noun that the “I” in the sentence is simulating is ‘la machine’. [TN]
 Apparently the French translation of Heidegger’s Verhalten is “comportement”, broadly meaning “behavior” or comportment (cf. Merleau-Ponty’s first book). The notion is to be taken in the sense of grammatical reflexivity (comporting-oneself). [TN]
 “Qui simule qui”, which could also read “who simulates whom,” referenced a few paragraphs above. [TN]
 This redescription of ‘uni-lateral Identity’ indicates why in the previous phrase “One (of) philosophy” only the first two words would be italicized, while the word philosophy is not. [TN]
Filed under: Translations Tagged: cognition, computers, immanence, Laruelle, machines, non-philosophy, performance, thought
Source: https://speculativeheresy.wordpress.com Translation of F. Laruelle’s “The Transcendental Computer: A Non-Philosophical Utopia”