Repetition as Machinic Appetition

Matteo Pasquinelli has posted a draft article titled Machinic Capitalism and Network Surplus Value: Towards a Political Economy of the Turing Machine. It is well worth a read and raises some very interesting points regarding the production of surplus value in the contemporary socio-technological juncture. Drawing on the work of Alquati, one of Pasquinelli’s more elegant points is thus:

Valorising information is then what enters the cybernetic machine and it is transformed into a sort of machinic knowledge. Specifically it is the numerical dimension of cybernetics that is able to encode workers’ knowledge into bits and consequently transform bits into numbers for economic planning. In other words, cybernetic code transforms information into value. […]
At the beginning of the industrial age capitalism was exploiting human bodies for their mechanical energy, but soon it was realised that the series of creative acts, measurements and decisions that workers constantly have to take is the most important value they produce. Alquati defines as information precisely all the innovative micro-decisions, which workers have to take along the production process, that give form to the product but also to the machinic apparatus.

I’ve also been interested in the question of value and the machinic for a while; admittedly, however, my clumsy forays are not as sophisticated as Pasquinelli’s work, which is genuinely very exciting. He invited feedback to his draft and I hope to offer some here in whatever meagre way I can contribute. I want to offer an alternative reading of the ‘assemblage’ from A Thousand Plateaus to that of the Delanda-esque inspired reading that Pasquinelli’s emphasises in his paper. Pasquinelli rightly indicates that the machinic as ‘productive’ as been reduced or even ‘neutralised’ of this dimension through the ‘relational paradigm’ of assemblage theory as forwarded by Manuel Delanda’s work (10). Without discounting Pasquinelli’s reading of Delanda’s somewhat de-weaponised version of the assemblage, I want to emphasise Guattari’s post-Lacanian reading of Deleuze’s work in Guattari’s essay Machine and Structure as indicating an alternative reading of the ‘assemblage’ that is at least congruent with Pasquinelli’s thesis.

I am working from the version of Guattari’s 1969 essay published in English in 1984 as a rough synthesis of Guattari’s works titled Molecular Revolution. In the second footnote of the early English translation of the essay Guattari draws on Deleuze’s work (below) and in the body copy notes that “in reality, a machine is inseparable from its structural articulations and, conversely, that each contingent structure is dominated (and this is what I want to demonstrate) by a system of machines, or at the very least by one logic machine” (111):

To adopt the categories suggested by Gilles Deleuze, structure, in the sense in which I am using it here, would relate to the generality characterised by a position of exchange or substitution of particularities, whereas the machine would relate to the order of repetition ‘as behaviour and viewpoint relative to a singularity that cannot be changed or replaced’ (D&R [Fr], 7). [In the 1994 Patton translation of the section I think is being quoted here — no other passage in the Patton translation comes close to capturing the meaning of Guattari’s point — it is translated thus: “[Repetition] is by nature transgression or exception, always revealing a singularity opposed to the particulars subsumed under laws, a universal opposed to the generalities which give rise to laws” (5).] Of Deleuze’s three minimum conditions determining structure in general, I shall retain only the first two:
1) There must be at least two heterogeneous series, one of which is defined as the signifier and the other as the signified.
2) Each of these series is made up of trms that exist only through their relationship with one another.
His third condition, ‘two heterogenous series converging upon a paradoxical element that acts so as to differentiate them’, relates, on the contrary, exclusively to the order of the machine (LoS [Fr], 63).

Guattari is clearer (or at least the translation is clearer) in the recent English translation and publication of The Machinic Unconscious, probably in part because he no longer had to participate in system of enunciation or ‘territory’ of structuralism and instead is operating within a properly Guattarian assemblage, where he writes:

The difference between Thom’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 [sic] “extracted” from the cosmos and history. Rather than abstract machines, perhaps it is preferably to speak of “machinic extracts” or deterritorialized and deterritorializing machines. […]
I will start with the idea that assemblages of flows and codes are first compared in relation to differentiations of form and structure, object and subject, and that the phenomena of formal interaction consitute only a particular case, that of a borderline case, within the machinic processes that work upon the assemblages before the substance-form coupling. (14, 15)

I want to suggest that the machinic, in as much as it is productive, is the concept that Deleuze and Guattari use to describe the milieu of repetition, in the specific singular or forceful mode of ‘repetition’ that, for example, makes and unmakes the relations of substitution of particularities and generalities contracted as habit or projected as the ‘social’ or as the conjunctive and disjunctive syntheses as described in Anti-Oedipus are the gathering and breakage of flows as ‘repetition’ moves across them.

Therefore, there is a problem when discussing the relation between the machinic and machines. Compare the above to the ‘actual’ machine that Guattari describes in Machine and Structure:

Initiation into a trade and becoming accepted as a skilled worker no longer takes place by way of institutions, or at least not those envisaged in such statements as ‘the skill has precedence over the machine’. With industrial capitalism, the spasmodic evolution of machinery keeps cutting across the existing hierarchy of skills.
[…]
As compared with the work done by machines, the work of human beings is nothing. This working at ‘nothing’, in the special sense in which people do it today, which tends more and more to be merely a response to a machine — pressing a red or black button to produce an effect programmed somewhere else — human work, in other words, is only the residue that has not yet been integrated into the work of the machine. (112, 113)

In other words, humans no longer repeat, even when their labour is repetitive; rather, repetition belongs to the machine as a milieu of the ‘machinic’. Both the worker and the machine belong to the machinic. The evolutionary character of the machine is machinic in the sense that ‘new’ evolutions assemble on yet another plane of consistency and produce new territories. These territories are sub-jacent to those that exist and have the effect of being inherently dissonant. The worker in the above is not a force of repetition. Guattari is restricting his analysis to the worker on the floor, and not the synthesised ‘workers’ of Pasquinelli that combines all non-machine human agents that are in some social relation to each other (in conflict or solidarity) as mediated by an actual machine. (So designers, engineers, technicians as well as the button pushers, etc.) In Anti-Oedipus, Deleuze and Guattari call this force of repetition — distributed across the knowledge worker as much as the labour worker and the machinic — the “surplus value of flux”.

Whitehead’s discussion of appetition provides a way to talk about both the tendential relations of futurity and the open, cutting edge of repetition as it is distributed across the machine and the worker (in Pasquellini’s broader sense). The appetiton of ‘repetition’ is immanent or self-posited (quasi-cause); or, following Derrida, it is ‘to-come’. It is only ‘discovered’ (or to use current parlance ‘innovated’) in retrospect, even if it is a retrospective projection of the imagined future into the present (i.e. ‘futurists’, ‘forward thinkers’, mythologised ‘Steve Jobs’, etc.). This is the work of capital: to extract the surplus value of the machinic from the force of repetition (in the specific sense of extraction of singularity, the con/disjunctive syntheses of flows and the redistribution of relations of particularity and generality through de/reterritorialization). Surplus value is extracted according to the “axiomatic of the world capitalist market,” following Deleuze and Guatari in Anti-Oedipus (233-235), that ever expands the immanent limits of capital itself.

The way I have interpreted Pasquellini’s essay — in the context of ‘surplus value of flux’ describing Deleuze and Guattari’s adaptation of something akin to ‘machinic repetition’ in the context of Anti-Oedipus — the appetition of the algorithmic machine is not this self-positing appetition of repetition and the milieu of the machinic. Rather, the algorithmic belongs to specific assemblages (or strata populated by apparatuses of capture, to use Deleuze and Guattari’s terminology) that have ‘captured’ the flux of repetition as surplus value (i.e. as ‘innovation’) and which (re)distributes the particular and the general in specific (‘actual’) ways. The appetition of the algorithmic machine is fundamentally inhuman while at the same time entirely the residue of human labour. What appetites does ‘Google’ have? What character of prehension belongs to an algorithm when it is ‘run’? I don’t know. Deleuze and Guattari developed the concept of the assemblage (machinic assemblages, but also the intertwined collective assemblages of enunciation) in A Thousand Plateaus as a way of grappling with the a-human territorializations immanent to machinic repetition: “In every respect, machinic assemblages effectuate the abstract machine insofar as it is developed on a plane of consistency or enveloped in a stratum (71). So I argue that accounting for the assemblages within which the surplus value of flux can be extracted through specific apparatuses of capture is essential, rather than being discounted. Pasquinelli already indicates the work of assemblages in his brief mentions of the socio-technological assemblages of Facebook, Google, etc. in his essay. The emphasis (and dismissal) of Delanda’s interpretation of the concept and development of ‘assemblage theory’ doesn’t necessarily foreclose, if repeated in different ways, productive use of the concept. If the algorithm could self-posit it’s appetites then it would belong to the “surplus value of flux” as a pure force of repetition and the properly ontological realm of the machinic, but otherwise, I want to suggest, it belongs to an assemblage.