Algorithms are a combination of logic and control (Kowalski, 1979). They describe the steps needed to complete a task (Goldschlager, Lister, 1988). They are sometimes thought of as discrete entities separate from the language they are programmed in or the way they are expressed. Details about how they are implemented fall outside of their existence as abstract objects, but connect them with the real world. In computation they are often found in software. Their instructional role aligns perfectly with the role of software in interfacing between humans and machines to solve problems, an idea proposed by Turing (Herken, 2013).
Even though they are abstract in nature, algorithms do bring about actions with all their attendant consequences. This suggests there is more to them than just being formal procedures with material outcomes. They are enmeshed within society and culture in myriad ways. Algorithms in computer art can be deterministic, producing the same result every time, or random (or at least pseudo-random) where a different result is produced each time the programme runs. Examples include generative art, cellular automata, fractal art and genetic algorithms.
Foucault’s concept of the statement is of relevance. He described the statement as a function of language’s existence which does not include it’s letters, words, grammar or even the meaning these words form (Foucault, 1969). The statement refers more to the in between domains or spaces in which language appears and forms subsequent relations. We can look at algorithms in a similar way. Whilst algorithms can be used to produce culture, they can also be used to enforce rules or laws, or collect data through surveillance of a population. Algorithms form part of the architecture of a modern panopticon (Foucault, 1975).
Algorithms could be seen as a way to communicate with machines, part of a machinic discourse. However this brings to mind the dualities of theory and practice, pure and applied, arguably dangerously ignoring the scope of the way algorithms operate. Instead of a purely abstract form, perhaps it is more helpful to look at algorithms and their relations to data, humans and machines as part of the infrastructure of a larger set of power relations.
Algorithmic identity (Cheney-Lippold, 2011), explored in artist James Bridle’s work “Algorithmic Citizenship”, is derived from the systems tracking us and the subsequent data collected for analysis and commercial exploitation. We are algorithmically sorted into digital categories of identity leading to a process of “software sorting” (Graham, 2005). Access to critical goods, services, life chances or mobility opportunities are determined by our social group or geographical area. Inequality is hard coded into the spaces we use to produce and access content, one example being the Prevent programme (Elgot, 2017).
One could go further and say the distinction between data and algorithms is purely formal and does not have relevance practically as both are required for outcomes to occur. The translation from one state to another which occurs through the algorithm is an incorporeal part of this ill-defined network of power-knowledge relations. With even only small events or changes in this network, it is possible for profound unintended outcomes to occur. Lessig said “Code is law” (Lessig, 1999), but all laws can be bent or broken. Not all logic is complete, there can be things which are unknowable or unpredictable. Within this space for infraction there is room for creativity.
Bridle, J (2015) “Algorithmic Citizenship” http://citizen-ex.com/citizenship
Cheney-Lippold, J. (2011) “A New Algorithmic Identity. Soft Biopolitics and the Modulation of Control.” Theory, Culture & Society, Vol. 28(6):164-181
Elgot, J. (June 2017) “Tim Farron warns of win for terrorists if web is made surveillance tool” The Guardian https://www.theguardian.com/politics/2017/jun/05/tim-farron-warns-of-win-for-terrorists-if-web-is-made-a-surveillance-tool
Foucault, M. (1969) “The Archaeology of Knowledge”
Foucault, M. (1975) “Discipline & Punish: The Birth of the Prison “
Graham, S. (2005) "Software-sorted geographies" Progress in Human Geography 29, 5, pp. 562–580
Goldschlager, G.; Lister, A. (1988) “Computer Science: A Modern Introduction” Prentice Hall International, second edition
Goffey, A.; Fuller, M. (Ed.) (2008) “Software Studies. A Lexicon” pg 15-20, The MIT Press
Herken, R. (2013) “The Universal Turing Machine: A Half-Century Survey”
Kowalski, R. (1979) “Algorithm = Logic + Control”
Lessig, L. (1999) “Code and Other Laws of Cyberspace”