New
Materialism

Perceiving algorithms has always been tricky. Bound up in recent times with ever pervasive web technologies and the proliferation of data, algorithms have often invited the same assumptions of immateriality levelled at all things digital. Indeed, simply understood as ‘instruction’, the algorithm appears even more ethereal than ‘data’ itself. However, obscurity has also been intentional. Algorithms are habitually concealed behind the slick visual facades of software, and the promises of abundant and horizontalised connectivity through social media. Where they are acknowledged, claims are often made about the revolutionary potential of algorithmic insights for science, economics, education, and entertainment. Yet, the assumption of objective and neutral results shrouds the underlying processes further, and functional understandings of algorithms are left to an emerging class of ‘data scientists’.

Critical algorithm studies are emerging, often with an agenda of exposing inner-workings and external influence, and have made the case for the algorithm’s increasing authority in culture (Striphas, 2015), social interaction (Gillespie, 2014), and politics (Tufekci, 2015). It is this tension - between a pressing need for critical examination and an inherent incomprehensibility and inscrutability (Ziewitz, 2015) – that marks the present condition of our attempts to perceive algorithms. For some, a mainstream and uncritical acceptance of algorithmic culture resembles a ‘computational theocracy’, where our relationship to algorithms ‘is not a material phenomenon so much as a devotional one’ (Bogost, 2015). Importantly then, both believers and critics need to avoid over-estimating or mythologizing the algorithm’s place as an all-powerful arbiter of social life, or indeed a foundational explanatory metaphor for the contemporary condition.

Nevertheless, algorithms present a crucially important site for research. A new materialist engagement with algorithms offers ways of drawing together procedural and cultural interpretations of the algorithm, and recognising the sociomaterial relations through which they operate:

Google is a monstrosity. It’s a confluence of physical, virtual, computational, and non-computational stuffs—electricity, data centers, servers, air conditioners, security guards, financial markets. (Bogost, 2015)

Thus the focus is on algorithmic assemblages, and their intra-acting relations, rather than straightforwardly on the input > process > output model of the algorithm-as-instruction.

Two directions are productive here. Firstly, the acknowledgement of the material dimensions of algorithmic processes. This involves an ecological view that encompasses database architecture, data storage centres and cloud infrastructures, as well as the complex spatial enactments brought into being through their algorithmic relations. Also implicated are the bodies, cars, and books made mobile by algorithmic processes, and whose mobility co-writes subsequent algorithm design and implementation. Dig deeper, and find the mining operations and rare minerals required to drive the computational turn. Sense the affective, where algorithms are deployed ‘to induce moods to make people more likely to consume their products’ (Tufecki, 2015, p. 212).

Secondly, recognition of a shift in the understanding of agency as no longer the exclusive possession of human subjects. Algorithms now not only predict best seller books[1], but also make death threats[2] and purchase illegal drugs on the dark web[3]. However, it is not enough to say that algorithms simply act and decide as human beings might. Rather, the operation of algorithms can be viewed as part of much broader and complex assemblages of action and decision, influencing human and algorithmic responses, but also being influenced themselves in the specific kinds of processes and calculations they undertake. A new materialist reading of the algorithmic shifts to a more-than-human register, recognising that most of the data flows that flicker across the earth take place between machines. Human action and decision is only part of this cognisphere (Hayles, 2006), in which algorithms already execute at imperceptible speeds[4].

[1] https://www.theatlantic.com/technology/archive/2016/09/bestseller-ometer/499256/
[2] https://www.theguardian.com/technology/2015/feb/12/randomly-generated-tweet-by-bot-investigation-dutch-police
[3] https://www.deepdotweb.com/2015/01/18/bot-drugs-deep-web/
[4] https://io9.gizmodo.com/a-new-digital-world-is-emerging-thats-too-fast-for-us-1286428447

Keywords: algorithm, data, digital, software, social media, more-then-human, cognisphere
Genealogies: cybernetics, computer science, computational processing, database, ‘data scientist’
Synonyms: instruction, code, calculation
Hypernyms: computer science, computational processing, big data, statistics
Hyponyms: data, code

References
Bogost, I. (2015). ‘‘The Cathedral of Computation.’’ The Atlantic, January 15. Accessed August 20, 2015. https://www.theatlantic.com/technology/archive/2015/01/the-cathedral-of-computation/384300/
Gillespie, T. (2014). The Relevance of Algorithms. In Media Technologies, ed. Tarleton Gillespie, Pablo Boczkowski, and Kirsten Foot. Cambridge, MA: MIT Press. pp. 167-194.
Hayles, N. K. (2006). Unfinished Work: From Cyborg to Cognisphere. Theory, Culture & Society, 23(7–8), 159–166. https://doi.org/10.1177/0263276406069229
Striphas, T. (2015). Algorithmic Culture. European Journal of Cultural Studies 18(4-5)
Tufekci, Z. (2015). “Algorithmic Harms beyond Facebook and Google: Emergent Challenges of Computational Agency” Colorado Technology Law Journal. v13 n2 https://ctlj.colorado.edu/wp-content/uploads/2015/08/Tufekci-final.pdf
Ziewitz, M. (2015). Governing Algorithms: Myth, Mess, and Methods. Science, Technology, & Human Values pp.1-14. DOI: 10.1177/0162243915608948