Algorithmic Thought: a review of Contagious Architecture by Luciana Parisi

From the start, in Contagious Architecture, the proposition is this: uncertainty and incomputability are intrinsic to computation. Looking for ways of accounting for a new digital space composed of ‘alien rule’ and ‘internal anomalies’ and of defining an algorithmic aesthetics proper to the contingencies, abstractions and potentialities of code forms the area of concern of this book.

In computer science, algorithms are habitually defined as fixed and often finite procedures of step-by-step instructions understood to produce something other than themselves; i.e. as time-based teleological structures that may interact with other things such as data, interfaces or data structures, with possible sensorial effects addressed to humans. According to Luciana Parisi, this approach refuses to acknowledge the reality of algorithms as also independent, both abstract and actual entities. As objects that seem to occupy the middle zone or the gap, between discreteness and infinity, abstraction and actuality, calculation and indeterminacy, algorithms – particularly, for Parisi, in the case study of digital architecture – reveal computation to potentially be a form of speculative thought in itself. At the same time, and according to their treatment in the book, they expose reason and logic to the novelty and creativity of the incomputable.

Incomputable quantities are at the core of this new algorithmic logic that the book strives to engineer, as discussed in Chapter One. The concept draws on the work of mathematician and computer scientist Gregory Chaitin, who places a real but ‘uncomputable’ number (Omega) at the core of his Algorithmic Information Theory. Omega defines an algorithmically incompressible real number located somewhere between 0 and 1 which at some point in its calculation forces the limited resources of any computer to come to a halt. Omega therefore suggests that there is no codified simplicity at the bottom of complexity. Simultaneously a real continuous number and something generated by discrete steps and manifest as discrete numbers, Omega seems to defy either/or categorizations. As Chaitin remarks, ‘the right way to think about each bit is that it’s not black or white, it’s not that it’s a 0 or a 1, it’s so delicately balanced, that it’s grey’ (2005, 23, his emphasis). The grey gap in between real numbers breeds digital potential, which escape the certainties of code, might not be sensed at all, and do not address cognition. Omega, thus, is used to challenge a discrete model of the universe as well as permit room for conceiving number outside the confines of finite, perfect, precise mathematical models. If, as Chaitin (2005) maintains, it is possible to have a number that has ‘no pattern or structure’ existing as ‘a string of unrelated digits’ that are ‘maximally unknowable’ and point to ‘an infinite series of facts that are…true for no reason’, then, Parisi proposes, it is time to acknowledge computation as a pure event of contingency; infused with randomness and elements that are incomprehensible by and independent of the human mind.

Hence, Contagious Architecture challenges the metacomputational approach of understanding algorithms according to a set of simple rules whose combination produces complex results, which is commonly at the heart of digital philosophy, for example, in Stephen Wolfram’s A New Kind of Science. It does so by unpacking an altogether different discrete philosophy that attempts to think its internal discrepancies and speculate on what algorithms are and what they can do. This revision of the digital involves an analysis of (first and second order) cybernetics and information theory, as well as their subsequent impact on debates about mediation, interactivity, cultural aesthetics, power, and control. First–order cybernetics refers to closed, self-sufficient systems that are observable from the outside and to causal processes, such as control, feedback, and adaption. Second-order cybernetics, which Parisi argues constitutes the dominant interactive paradigm of capitalism today, revolves around the idea that the observer is also part of the system and concerns notions of reflexivity, self-organisation, autopoiesis, the contingency of environmental factors, and the indeterminacy of living systems. The book follows the conjecture that this second interactive model, which is also at the core of digital architectural processes, does not quite reach its full potential and deliver what it had promised: that is, a new organisation of data and conception of the incomputable (chance, randomness, and complexity) as integral to computation. By building on the algorithmic production of infinity in Chaitin’s Omega, Parisi is able to extend the second level – in all its progressive solutions to first-order problems – towards a speculative third-order. The latter contents that a) logic and reason are becoming aesthetic operations defined by algorithmic prehension, and b) incomputability and (a new form of) algorithmic entropy (i.e. randomness) should be understood as intrinsic to any computation.

Both the speculative drive of the project and its mission to address the question of the digital, in its discrete, undivided, and united nature, emerge from an original engagement with Alfred North Whitehead’s philosophical and scientific writings. More specifically, Parisi turns mainly to his “philosophy of extension”, “the function of reason”, and “the principle of relativity”, where discontinuity is treated as primary and immanent whilst the ‘overall process’ comes second. For instance, Parisi quotes Whitehead, for whom, against pure idealism, the function of reason is “to constitute, emphasize, and criticize the final causes and strength of aims directed towards them”, which she interprets to mean that speculative reason serves “to unlock new possibilities within the order of things” (2013: 73). As Chapter One outlines, unlike object oriented philosophies – such as Graham Harman’s that refreshingly asks “how objects could be granted an autonomous existence that would not prevent them from entering true relationships” (48) – which do not completely do away with the notion of substance, Whiteheadian theoretical devices can enable a definition of algorithmic objects that “accounts for the presence of indetermination in quantities” (55). Parisi’s Whiteheadian treatment of the digital problem proposes that algorithms are actual, undivided, discrete entities containing infinity and imbued with randomness. Leaving behind the topological schema that presumes algorithms to be subjected to continuous, sequential order results and evolving in time,¹ the book turns instead to parametricism and the mereotopological order of events.

In Chapter Two, Parisi looks to Whitehead’s mereotopological approach , i.e. atomic spatiotemporalities that connect to one another but cannot be summed up to a whole, to oppose subsuming discreteness to continuity. Mereotopology suggests that spatiotemporal entities do not pre-exist but are purely the outcome of prehensions, crucially offering an alternative model of whole-to-parts connection. In Whitehead, prehensions define what an actual occasion (event) is and how it relates to others. In the book, as the glossary states, prehension is “the process by which an actual entity confronts infinite data … and by which it thereby invests and reprograms the actual field of potentiality” (266). From this standpoint, prehension can help explain why actual objects are always incomplete and “how indivisible and discrete unities can exist in the infinity of relations with other events” (60). More than this, prehension is here equated to contagion, as Parisi explains that to prehend is to be infected with infinite varieties of quantities but without, in turn, being able to change them. Therefore algorithms are actual objects that prehend both the formal system into which they belong but also the external data they retreat. This view is very different from the topological standpoint of parametricism, the current global style of fluid architecture, which assumes events as deriving from a relational continuity between infinitesimal points of contingencies that lie outside the program.

By plugging into the mereotopological schema, which unlike topology denies any reciprocity between control and events, Chapter Two argues that parametricism can expose the increasing autonomy of algorithmic entities, quantities, and parameters from human thought. With the help of cutting edge architecture, such as found in the practice of François Roche, this part of the book ponders how parametric quantities are discrete entities that not only select data, as part of the computational program into which they are scripted, but might also be infected by data that they are not able to compute. “Instead of being a continuous flow of data, such as a topological binding of many actualities into one stream of ceaseless variation, the incomputable … is an infinite series of discrete yet incomplete data that immanently ingresses and becomes uniquely arranged into algorithmic sets, in which these data acquire togetherness and continuity” (170).

Parisi’s theorisation of the incomputable quantities underlying algorithmic processing, leads to matters beyond mathematical and technological specificity. It suggests that new aesthetic forms are developing in the hypercapitalist sphere affecting and affected by current design and computational practices and prompting a reconsideration of the questions of power and control. The vigour of the book truly comes forth in places where Parisi puts forward new concepts and uses them as probes to get to somewhere new; such as with her term soft(ware) thought in Chapter Three, which she conceives as a mode that is completely autonomous from cognition and perception. In this part, Contagious Architecture rounds up its larger intent of mapping computation – through its centrality to digital architecture – as a speculative operation traversed by infinities. If digital design can be seen as an instance of soft thought, it can also help demonstrate that this mode of thought is actually a form of immanent experience, not reducible to new spatiotemporal phenomenological experiences. “Rather, soft thought is experience defined by the algorithmic prehension of infinities … [it is an] event … a lived abstraction that has not been fully axiomatized (a quasi-empirical or quasi-formal computation)” (175). Juxtaposing the position that algorithmic thought is reducible to the neural structure of the brain or to the interactions between brain and world (neuroarchitecture/ neurophenomenology/ neurocomputation), Chapter Three upholds that soft thought expresses “the immanence of the incomputable in thought” (255); made of transitions between actual entities given in pure experience (Whitehead vis-à-vis William James). This final part of the book sketches out what might take a life of its own in Parisi’s future work, that is, an analysis of the relations between reason and automation. As she indicates, soft thought threatens the assumption that computation is just another function of reason and that algorithms are reducible to a human mind and sensorimotor system.

Contagious Architecture is an investigation into developing a nonanthropocentric mode of thought and a call to shift focus from essences and qualities to peripheral zones of attention, non-direct cognition & emotion, and immanent rather than directly lived experiences of spacetime. If, as it is now evident, this is the age of computational media infecting all modes of cultural expression, what Parisi calls ‘programming culture’, then there is a pressing need for a proper engagement with a new paradigm of speculative thought: one that is unapologetically asocial and ahuman. The book’s aesthetic view of algorithms challenges the equation of digital architecture to the logic of neo-liberal capitalism, suggesting a profound rupture in the rationalisation and efficiency of the latter through its enforced reliance on abstraction and indeterminacy. It is part of a wider attempt to overturn what is meant by the digital and to challenge its casual associations with interactivity, communication, and control, as inherited from the first wave of cybernetics and still haunting digital culture, art, and theory today. Contagious Architecture tells us there is room for a more dynamic engagement with algorithms, beyond a habitual analysis of the possibilities and probabilities of computation that says nothing about data’s contingencies and potential. In so doing, it brings forth the potency of an unlived reality inseparable from computation and lurking in the shadows of the everyday.

This intention is shared by a current wave of media theorists and a body of work that, on the surface, seems to undertake an impossible task: human thought acquiring nonhuman perspective. But why now? What enables, probes, and unleashes this urgency to rethink the digital by speculating that we do not even know what it can do? What are the conditions that allow an intervention into the algorithmic unknown that perhaps were not there before – for example, when Deleuze was referring to the digital’s weakness in accessing the virtual in his ‘Analogy’ passage in the Francis Bacon book; or when Massumi identified that a power to effectuate change was missing from the digital in his essay ‘On the Superiority of the Analogue’. It seems conceivable that this ‘now’ cuts across an infinity of spatiotemporalities, not purely restricted to existing capacities in technology and new media experimentation; or to the occasion of viral abstract thought slowly contaminating the arts and humanities and blurring the boundaries of disciplines; or even as an unstated event in itself, whereby certain works directly engaging with the weird speculative dimensions of computation are drawn together to collectively tell the untold tales of theory. Contagious Architectures belongs to this realm of risky thought, dissecting the ways in which algorithms might exceed determination, and, in so doing, daring to explore a capacity in thinking beyond or without decision.

Notes