Unknown Unknowns: Automaticity & Agency

Agency is the extension of the actant into a social relation, attempting to implement their desires and enact commitments. In this extension, the values and commitments of actants come up against those of others. Commitment as extension moves being from a passive subjectivity (caught within the flows of intercourse that formulate an advanced sentience) toward a referential agency (referencing the understanding of these flows as a form of sapience). A move from the limited individuation of a self-interested perspectival mode[1] toward a greater collective understanding with attendant conflicts and negotiations between different agents in these collective assemblages. This extends beyond the simple relations of human formations, into wider assemblages or paradigms as Latour describes them[2]. In paradigms, we see a wide field of potential actors producing relations and oppositions. Latour uses the story of a hotel manager, who in encouraging his guests to leave their room keys at the front desk before leaving the hotel, attempts various technological and social methods, from warning signs and reminders to changing the design of the keys themselves, adding a small weight which makes the keys bulkier and less desirable to carry around. These are the syntagms that are built upward in paradigmatic extensions. In these continual extensions, we see the emergence of an agential field, including not just the hotel manager but the lay out of the hotel (with added signs and reminders) and the construction and design of the key itself, producing a particular technical assemblage where the manager’s agency is mixed with a technicity of control.

This description has similarities with Negarestani’s conceptualisation of the inhuman as the affectation of constructed (or augmented) reason. In the formation of norms and agential properties, “reason has its roots in social construction, in communal assessment, and in the manipulability of conditionals embedded in modes of inference. It is social partly because it is deeply connected to the origin and function of language as a de-privatizing, communal, and stabilizing space of organization”[3]. Reason is the outward condition of relationality in producing implicit norms. It sits in the middle of the natural and social, being the middle ground between unconditional acceptance and unlimited freedom as it acknowledges laws.

Reason in this autonomous capacity for acknowledgement becomes an actant in extensive fields of social action. It develops an automatic discursivity, focused on the adaptation of social laws to human discretion, creating that middle ground mentioned. It exists outside of the human faculty as a semi-autonomous intelligence, implying “the automation of reason, since the autonomy of practices, which is the marker of sapience, suggests the automation of discursive practices by virtue of their algorithmic decomposability into nondiscursive practices”[4]. This automaticity is both its own agency, in that it extends particular values (the production of subconscious intelligence through normalisation and the extension of particular conceptions of the human) and is intimately related to human agency for the same reasons.

Negarestani then posits this automaticity as an engineering epistemology, “an upgradable armamentarium of heuristics that is particularly attentive to the distinct roles and requirements of different levels and hierarchies. It employs lower-level entities and mechanisms to guide and enhance construction on upper levels. It also utilizes upper-level variables and robust processes to correct lower-level structural and functional hierarchies, but also to renormalize their space of possibilities so as to actualize their constructive potentials, yielding the observables and manipulation conditionals necessary for further construction”[5]. This hierarchical concept has affinities with a Weberian bureaucracy, in that the extent of roles or subsystems is limited by their place within the wider system. The 4I feedback loop (intuition, interpretation, integration and institutionalisation) is instructive in describing the linear dynamics of this knowledge structure. Going from the initial condition of social interactions within groups, the interrelations of human and reason first create conflicts between learned and technical knowledges, or between past and future as the act of intuiting means working through existing problems, thereby questioning a foreground of existing (or learned) knowledge. This moves up toward interpretation as these knowledges begin developing alternate practices, variegating the existing field of practice as well as knowledge. Integration then occurs through reuse and scalability, as new methods outcompete old. Finally institutionalisation is the completion of the discursive circle, as the products of human-reason relations produce new methodologies of action which are sublimated, intuited and interpreted by a wider field of actants.

This explication is revealing of a systematic ontology structured through integrated hierarchies that feedback to each other, with intuitions feeding back into higher level structures which in turn send communications back down the chain so that the field of actants is integrated into institutional norms. This has similarities to the constraints-based model of institutionalisation in that both presuppose the production of social facts that act as constraining devices on actors within particular institutions, with Negarestani seeing the constraints as produced through an autonomous reason that automates these discursive flows into non-discursive conditions of understanding the self.

And it contains the same problems as the constraints-based model in conceiving of a rigorised agency consisting of primary negotiation, secondary interpretations and a sublimated system of learning through institutions. In developing an engineering epistemology, it reifies structures of facticity that are rationalised. By processing social facts through reason, a stripped back form of agency is produced akin to Cudworth & Hobden’s[6] reproductive agency combined with affective agency, with actions controlled with structural hierarchies, and the capacity for influence dependent upon one’s level in this hierarchy. This is a formalised schema that, while containing contingency, presents an expansive concept of intelligence that can internalise contingencies or black swans.

The capacity for informal practices outside these structural constraints, or for opposition via exit, appear limited and even potentially extensible from the structures themselves. In much the way institutional cartography[7] attempts to construct an incontestable geography, an engineering epistemology appears to limit facticity to a discursive duality of competitive success or anti-reason (which is discarded). It also produces limits to the field of potential actants in conceiving the human and reason as discursive intelligences that continually expand outwards, reproducing a teleology. “It is the acknowledging, error-tolerant, revisionary dimension of ought—as opposed to the impulsive diktat of a natural law—that presents ought as a vector of construction capable of turning contingently posited natural necessities into the manipulable variables required for construction”[8]. This organisational processualisation ignores the other actants that Latour notes, that are integral to the development and evolution of networks of social action i.e. technologies, ideologies, intercursive flows, etc. These are channelled through a macroscopic institutionalism and a collective human agency.

The ability for things to run beyond agential control, or to develop autonomous agencies, is limited in such structural constraints. Instead a technocratic subject is reproduced, where the potential of human agency strained through institutional norms is given precedence in relation to the direction of events. Negarestani’s agential configuration would suggest that these events are constructable, in that through the directionality of reason future concepts can be brought into the present to reconfigure the past due to the requirements of adaptability in the face of ossification. Rather than entropy being a limit to the capacities for structural knowledge production, there is the capacity for convergence through the production of new norms and facts.

Going back to Latour, the interactions of actor-networks meet constant limits through anti-programmatic actors and systems. Rather than convergence toward evolutionary development, the growth of paradigms is reliant upon negotiation with those affected and with the possibility for syntagms to be reformed, scaled or opposed. This isn’t a world of reason as such, but of forces of will interacting with each other through conflict or cooperation. Programmatic extensions like the use of technology in Latour’s hotel are reliant on the construction of shared desires or shared limitations, with the capacity for anti-programmatic action (that isn’t centred on reason) always in the foreground of innovations. While reason can extend itself through the production of (in)human cognitive capabilities and structural transformations through feedback, the ability for alternate desires and uncontrollable agencies will always remain.

“Yet just as metaphysics is created in the (abstract) image of ‘man,’ so too this ‘deception’ is made to prevail over what is in fact most alienating to ‘man’ in the Anthropocene: a world that is not the object of Reason, but which stands fully in place of Reason; a world not subsumed by technology, but constitutive of it; a world not alienated by ‘human’ forces that have obtained, in its termination, an irrefutable ascendancy & mastery (‘the destiny of the Earth’), but which is that alienation in all its radical ambivalence”[9]. This primary alienation is itself constitutive of the structural relationality seen in reproductive and affective agencies, as both constrain and sublimate desires through their vectors. But in doing this we are not just talking about structural constraint, but structural agency in that these structures develop their own epistemologies beyond simple human control.

Marx’s general intellect begins to show this in the production of an industrial ontology, a force of will over the constituent parts, creating an assemblage of individual and class agencies into an autonomous productive force. “The foundations of Marx’s ‘general intellect’ ultimately reside in those operations of entropy in which the so called forces of nature themselves originate (as ‘man’s inorganic body’) & in which the dynamic of ‘alienation’ evolves towards a consciousness & a production of subjectivities that is not modelled on the human but produces it”[10]. Whether emancipatory or totalising, this production of an intellectualism is not purely the product of reason or of human intentions, but an extension of productive forces developing their own agencies. These of course interact with human agencies, producing alternate means of production through the interactions of capital (as force of will) and individual components (or oppositions). Thus the productive developments of Taylorism, Fordism or Post-Fordism are ontological logics that control the power of constituent parts, producing subjectivities. A plasticity of the object-subject distinction develops as productive forces produce their own agential dynamics, much as technologies transform the relations of the hotel from something between manager and guest to something involving multiple branches and rhizomes of potential interaction, exponentially multiplying the relations of paradigms with programmatic actions (or subversions).

Heidegger’s linguistics of technology shows a similar agential force in the way uncontrollable dynamics can evolve from the productions of reason. Techne as a force for the development of technicity presents a force among the quanta in conflict. The introduction and expansion of technology produces ontological and epistemological variations of its users, dividuating them as Raunig describes by splitting and segmenting elements of their personality through different emphases. Different linguistics emerge that move beyond reason or human understanding. Decentralised autonomous organisations and blockchains are the most obvious examples of this dynamic, with autonomous codifications and internal systems of trustlessness (“a truth machine”[11]) that move beyond individual or collective control. “Heidegger observes that because of technology, ‘all distances in time and space are shrinking’ and ‘yet the hasty setting aside of all distances brings no nearness; for nearness does not consist in a small amount of distance’”[12]. In the development of a new chronicity, a new subject is being honed which moves beyond the bounds of reason as it doesn’t just require agential adaptation but also instantiates conflict with existing perspectives and knowledge structures. A uchronia[13] of increasing speed, of perceptivity, relationality and personality as technologies warp understandings of being, interrelating these agencies much as reason interrelates with the human. Technological autonomy is an extension of an alien agency, conflicting with established lifeways. Agricultural innovation is indicative of this dynamic, moving from smallholdings, seasonality and crop diversity to intensive methods, year-round production, increasing acreage and monoculture (as through the Green Revolution). This is transformative and dislocating of the farmer as a class subject and concept of being.

Even reason shows similar dynamics to those of technicity in producing after-human agencies and wills. As I mentioned, the automaticity of reason’s extensions are similar to a technocratic political subject. Looking at existing variations of technocratic politics, we see the production of new subjectivities and relations of power. A commitological position is the means of political expression within the majority of bureaucratic agencies (like the Federal Reserve or the European Union). Nested hierarchies with an overarching competence that defines their agential dynamics. The anti-democratic nature of these formations move them beyond the control of the political animal, instead being enmeshed in relations of expertise through transnational class structures. This is a transformation of the governing relations informing democratic nations from the post-war era, as the relations of party structures intimately linked with industrial structures and policies have given way to a post-political constitutionalism of expansive legal codification, whether through voluntary guidelines and terms of service or through greater litigious power through the increasing interventions of the US Supreme Court and the ECJ. This extension of the technocratic subject reifies reason and rationality as the mechanisms for political action, relying on the expertise of jurists, economists and other epistemic communities that go beyond any popular control, creating alternate languages of expression removed from common vernacular.

The development of unknown unknowns move beyond the purposive capacities of reason or human agency, as autonomous systems with their own causal determinations and linguistics (and symbology) extend their own control, conflicting with human perspectives and making reason a contestable field for differing interpretations (rather than a post-political factuality). This can be conceived as the difference between logistics and logisticality. “Logistics is in hot pursuit of the general intellect in its most concrete form, that is its potential form, its informality, when any time and any space and any thing could happen, could be the next form, the new abstraction”[14]. It is the extension of a rationalised system of production, which sublimates its constituent agencies and produces a collective intelligence as an assemblage. Logisticality is the incalculable outside of logistical perception. It is the production of subjectivity within the relations of objects, information and other intercursive flows under an institutional structure, emerging as an alter-subject removed from established hierarchies which attempt to emplace the human within a particular epistemology (which then attempts to structure their being). Autonomous structures (including Negarestani’s reason) move beyond their originary conditions, developing alternate agencies that undermine or move beyond logistical control. Logisticality will always flow on the back of the entropy of actor-network relations, as actors constantly negotiate and alter the paradigm.

[1] https://vividness.live/developing-ethical-social-and-cognitive-competence

[2] Bruno Latour, Technology is Society Made Durable

[3] https://www.e-flux.com/journal/53/59893/the-labor-of-the-inhuman-part-ii-the-inhuman/

[4] https://www.e-flux.com/journal/53/59893/the-labor-of-the-inhuman-part-ii-the-inhuman/

[5] https://www.e-flux.com/journal/53/59893/the-labor-of-the-inhuman-part-ii-the-inhuman/

[6] Erika Cudworth & Stephen Hobden, Of Parts and Wholes: International Relations Beyond the Human

[7] https://thelibertarianideal.com/2021/02/15/institutional-oceanography/

[8] https://www.e-flux.com/journal/53/59893/the-labor-of-the-inhuman-part-ii-the-inhuman/

[9] https://alienistmanifesto.wordpress.com/2020/04/14/entropology/

[10] https://alienistmanifesto.wordpress.com/2020/04/14/entropology/

[11] https://cybertrophic.wordpress.com/2021/02/13/the-engineer-the-artist-negarastani-vs-land/

[12] https://www.thenewatlantis.com/publications/understanding-heidegger-on-technology