Introduction: Logos or Abstract Machines? (part 1)

Felix Guattari 

Does the unconscious still have something to say to us? We have saddled it with so much that it seems to have resolved to keep silent. For a long time it was believed to be possible to interpret its messages. A whole corporation of specialists worked away at this task. Nevertheless, the results were hardly worthwhile, for it seems likely they have all gone astray. Would the unconscious definitively speak an untranslatable language? It's quite possible. It would be necessary to start again from the beginning. First of all, what is this unconscious really? Is it a magical world hidden in who knows which fold of the brain? An internal mini-cinema specialized in child pornography or the projection of fixed archetypal plans? The new psychoanalysts have worked out more purified and better asepticized ideal models than the older ones: they now propose a structural unconscious emptied of all the old Freudian or Jungian folklore with its interpretative grids, psycho-sexual stages, and dramas copied from antiquity ... According to them, the unconscious would be "structured like a language." Yet, it goes without saying, not like everyday language, but like a mathematical language. For example, Jacques Lacan currently speaks about the "mathemes" of the unconscious ...

We have the unconscious we deserve! And I must acknowledge that the structuralist psychoanalyses are even less appropriate in my view than the Freudians, Jungians, or Reichians. I would see the unconscious instead as something that we drag around with ourselves both in our gestures and daily objects, as well as on TV, that is part of the zeitgeist, and even, and perhaps especially, in our day-to-day problems. (I am thinking, for example, of the question of "the society we choose to live in" that always resurfaces around the time of each electoral campaign.) Thus, the unconscious works inside individuals in their manner of perceiving the world and living their body, territory, and sex, as well as inside the couple, the family, school, neighborhood, factories, stadiums, and universities... In other words, not simply an unconscious of the specialists of the unconscious, not simply an unconscious crystallized in the past, congealed in an institutionalized discourse, but, on the contrary, an unconscious turned towards the future whose screen would be none other than the possible itself, the possible as hypersensitive to language, but also the possible hypersensitive to touch, hypersensitive to the socius, hypersensitive to the cosmos ... Then why stick this label of "machinic unconscious" onto it? Simply to stress that it is populated not only with images and words, but also with all kinds of machinisms that lead it to produce and reproduce these images and words.

We are accustomed to thinking of material and social facts in terms of genealogies, archaeological residues, and dialectical progress or in terms of decline, degeneration, and rising entropy ... Time goes on toward better days or plunges blindly toward unimaginable catastrophes; unless it simply stares to vegetate indefinitely. We can bypass these dilemmas by refusing any sort of causalist or finalist extrapolation and by strictly limiting the object of research to structural relations or systemic balances. But no matter how one goes about it, the past remains heavy, cooled down, and the future seems largely mortgaged by a present closing in on it from all sides. To think time against the grain, to imagine that what came "after" can modify what was "before" or that changing the past at the root can ​transform a current state of affairs: what madness! A return to magical thought! It is pure science fiction, and yet ... 

In my view, there is nothing absurd about attempting to explore these interactions, which I would also qualify as "machinic," without initially specifying their material and/or semiotic nature. Neither transcendent Platonic idea, nor Aristotelian form adjacent to an amorphous matter, these abstract deterritorialized interactions, or, more briefly, these abstract machines traverse various levels of reality and establish and demolish stratifications. Abstract machines cling not to a single universal time but to a trans-spacial and trans-temporal plant of consistency which affects through them a relative coefficient of existence. Consequently, their "appearance" in reality can no longer claim to be given all in one piece: it is negotiated on the basis of quanta of possibles. The coordinates of existence function like so many space-time and subjective coordinates and are established on the basis of assemblages which are in constant interaction and incessantly engaged in processes of deterritorialization and singularization causing them to be decentralized in comparison to one another while assigning them "territories of replacement" in spaces of coding. This is why I shall oppose territories and lands to machinic territorialities. By distinguishing them from set logic, a "'machinism" of the assemblage will only recognize relative identities and trajectories. It is only on a "normal" human-scale-i.e. that which pertains neither to madness, childhood, nor art- that Being and Tune will seem to thicken and coagulate to a point of no return. Having considered things from the angle of machinic time and the plane of consistency, everything will take on a new light: causalities will no longer function in a single direction, and it will no longer be allowed for us to affirm that "everything is a foregone conclusion."

Following Rene Thom, it even seems possible one can "take back one's throws," since, according to this author, the logos of the biological species would be able to operate a sort of "smoothing of time" in the direction of both the past and future. Due to the definition of this logoi and so long as "space-time figures, as well as their variations, conform to a principle excluding discontinuities and angles," the phenomena that refer to them would be able to influence their predecessors and their successors. Here as well, all in all, machines become independent of their immediate manifestations while "smoothing" a plane of consistency that authorizes every intersection possible! And yet, this logoi inspire only relative confidence in me. I fear that they merely have an irresistible inclination to escape from the physico-biological world in order to rejoin the mathematical universe of their origin! What particularly worries me is that they can only be f.tctored in, as Rene Thorn explains, so that the most abstract are arranged with the most elementary and the most concrete with the most complex. This simple fact seems to condemn them to definitively faU to maintain their hold on reality. The difference between Thorn's logoi and abstract machines, such as I conceive them, stems from the fact that the former are simply carrying abstraction, whereas the latter in addition convey singularity points "extracted" from the cosmos and history. Rather than abstract machines, perhaps it is preferable to speak of "machinic extracts" or deterritorialized and deterritorializing machines. In any event, I consider that they should not be comparable to entities attached once and for all to a universe of forms and general formulas. By preserving the expression "abstract machine" in spite of its ambiguity, it is the very idea of abstract universals that I aim to dispute. Abstraction can only result from machines and assemblages of concrete enunciations. And since there is no general assemblage that overhangs all of them, every time we encounter a universal enunciation, it will be necessary to determine the particular nature of its enunciative assemblage and analyze the operation of power that leads it to lay claim to such a universality.

The ideal of order-the systematic formalization of every mode of expression, the control over semiological flows, and the repression of the lines of flight and lines of dissidence-that dominates university research and the practical fields of the social sciences can never be completely attained, primarily because it is the stake of political and micro-political power struggles, but, perhaps more basically because, as we shall see, languages drift in all parts. Scientific formalization, fortunately, does not make an exception to this rule. The exhaustive dichotic analysis, binarism reduction, and radical "digitalization" of every semiotic practice, whose model has been elaborated by information theory, seems to function today {in league with behaviorism and Pavlovian theories, with which it also has certain affinities} as a sort of instrument of contention in the field of linguistics and the social sciences. We believe that such a method could in principle be applied to any type of social phenomenon. And if we manage to implement it through some sort of sleight of hand, we are then convinced to have grasped the essence of the phenomenon in question: we can stop and pass on to something else. Unless, while pushing things to the extreme, we come from that position to no longer consider any event except in terms of its probability of occurrence, and then, in the name of the sacrosanct second principle of thermodynamics, to proclaim that everything must tend towards a state of equilibrium or that every structural phenomenon must necessarily evolve towards a reduction of tensions and disorder. A few universal principles hangover contingencies and singularities, precisely with respect to probabilizing events on a diachronic axis and structuralizing them on a synchronic axis: this is what the ambition of the various structuralist schools is reduced too! In fact, I believe that this kind of operation always turns up in order to "sweep under the rug" the socio-machinic assemblages which are ultimately the only effective products of rupture and innovation in the semiotic fields that interest us here. Chance ​and structure are the two greatest enemies of freedom. They induce the same conservative ideal of the general axiomatization of the sciences that has invaded their field since the end of the 19th century. And since they have furthermore become inseparable from the philosophical tradition as a pure subject of knowledge inaccessible to historical transformations, they return us very quickly to the meddlesome and sclerotic discourse of epistemology. It is always the same juggling act: through the promotion of a transcendent order founded upon the allegedly universal nature of the signifying articulations of certain enunciations--the Cogito, mathematical and scientific laws, etc ... -one endeavor to guarantee certain types of formations of power, simultaneously consolidating the social status and the imaginary security of its pundits and scribes in the fields of ideology and science. 

​Two attitudes or two politics are possible with regard to form: a formalist position that begins with transcendent universal forms cut off from history and which are "embodied" in semiological substances, and a position that begins with social formations and material assemblages in order to extract some (to abstract some) of the semiotic components and abstract machines from the cosmic and human history that offers them. With this second path, certain "accidental" conjunctions between "natural" encodings and sign machines will affirm themselves, will "make the law," during a given period. However, it will be impossible to consider them independency of the assemblages that constitute the nucleus of their enunciation. It is not a question, as one could be tempted to say, of a re-enunciation. Indeed, there is no meta-language here. The collective assemblage of enunciation speaks "on the same level as states of affairs, states of facts. and subjective states. There is not, on the one hand, a subject that speaks in the "void" and, on the other hand, an object that would be spoken in the "plenum." The void and the plenum are "engineered" by the same deterritorialization effect. ​Connections are only possible at the point where abstract mechanizations and concrete, dated, and situated assemblages enable a connection to their deterritorialization. Also, assemblages arc not delivered randomly to the axiomatic of universals: the only "law" they uphold is a general movement of deterritorialization. The axiomatic returns to the assemblage more deterritorialized in order to solve the impasse of previous systems of enunciation and untie the stratifications of the machinic assemblages that correspond to them. Such a "law" does not imply a pre-established order, a necessary harmony, or a systemic universal of anything.

​Rene Thorn, who knew how to denounce the "dream of information theory" with humor, or rather the dream of those whose hopes depend upon a set of formal systems and morpho-genesis,' perhaps did not come to the end of his intuition. Does he not lend himself to the brunt of his own criticism when he sets out in search of a system of algorithms that would be able to give an account of every morphogenetic change, of every "catastrophe" capable of affecting an assemblage? 

He rightly considers that the "abstract logoi," immigrants of the physical and biological world, never stop "invading" the cerebral world. But there are many other continents from which such "invasions" develop, beginning with the world of socio-economic assemblages and that of the mass media. According to him, every interaction is brought back to the phenomena of formal resonance in the last analysis (page [200]). On the contrary, I will start with the idea that assemblages of flows and codes arc first compared in relation to differentiation of form and structure, object and subject, and that the phenomena of formal interaction constitute only a particular case, that of a borderline case, within the machinic processes that work upon the assemblages before the substance-form coupling.

​Abstract machines do not function like a coding system stacked on from the "outside" on the existing stratifications. Within the framework of the general movement of deterritorialization that I evoked earlier, they constitute a sort of transformational matter, what I call an "optional subject" -composed of the crystals of the possible which catalyze connections, destratifications, and reterritorializations both in the living and inanimate world. In short, abstract mechanisms emphasize the fact that deterritorialization in all its forms "precedes" the existence of the strata and the territories. Nor being "realizable" in a purely logical space but only through contingent machinic manifestations, they never involve simple combinations; they always imply an assemblage of components irreducible to a formal description. "Descending" from the pragmatic fields to the assemblages, from the assemblages to the components, and then from the components to the matters of expression, we shall see that we will not necessarily pass from the complex to the simple. We will never be able to establish a final systemic hierarchy between the elementary and the compound. Under certain conditions, the elementary can always make new potentialities emerge or make them proliferate and include the remainders within the assemblages to which it is related. Also, rather than starting with the elementary, which is likely to be merely a lure, the analysis will attempt to never simplify or reduce what seems preferable to call a molecular level. Machinic molecules may carry the keys of encoding that lead to the most differentiated assemblages. Moreover, the scope of the "most complex" generally seems to depend upon the fact that these molecular machinisms are more deterritorialized and more abstract. 

​​from the book: THE MACHINIC  UNCONSCIOUS (ESSAYS IN SCHIZOANALYSIS) by Felix  Guattari

​Translated by Taylor Adkins

to be continued