Computational Thinking Revisited

Computational thinking was originally framed as an educational and largely entrepreneurial skill set. Jeannette M. Wing defines it as “an approach to solving problems, designing systems and understanding human behavior that draws on concepts fundamental for computers.”^[1] Wing intended this set of thinking through computers to permeate educational facilities, from basic school to university, fostering a new generation able to work within computer environments.^[2] There is however a tension in her understanding of computing as “the automation of our abstractions” – a fundamental component of computational thinking since Aho and Ullmann’s Foundations of Computer Science (1992) – and her call to “define the right abstraction.”^[3] This critical distance from what the machine can do to what the machine should do has been a central issue in critiques of computational thinking during the last decade.^[4] If the user automates her abstractions based on what the machine can do, does she not lose sight of what the machine ought to be doing for the benefit of a wider ecology of relations? If goal orientation is involutional, the critique goes, we stand to lose much of what machines could be doing.

Although in her 2008 essay Wing does pose a few remarks on how “our abstraction tend to be richer and more complex [in computer sciences] than those in the mathematical and physical sciences,” she does not reflect on the ways in which abstracted machine code can structure the horizon of affect and potentiality. Thinking through machines is, in Wing’s take, commensurate with the entrepreneurial necessity for the industrial and economic sector,^[5] and she does not consider the work done within contemporary critical theory devoted to addressing the problems of governance and computation. In other words, she does not consider the fact that computational thinking could in fact circumscribe critical and creative thinking, insofar as it brackets itself as a rationale “in which the terms of engagement (the epistemological scene, so to speak) are pre-comprehended or predetermined.”^[6]

A nuanced and invested critique of computational thinking was offered by Steve Easterbrook in his essay “From Computational Thinking to Systems Thinking,”^[7] where he states that “since the concept was introduced, there has been remarkably little critical thinking about computational thinking.” Easterbrook regards “computational thinking as a major limiting factor” and proposes “to supplement it with systems thinking.” (emphasis in original) Because computational thinking is “inherently reductionist,”^[8] it needs to be coupled with a broader spectrum of analysis which might afford a more distant reading focusing on what ought to be done with the current affordances of transnational, quasi-planetary digital computation. 

The machine and the human have become ever more integrated within a material and virtual mesh woven of aesthetic interfaces and software. Where Benjamin Bratton has taken the structuring logic of computational assemblages and has developed a whole theory of the Stack,^[9] the stack as a fundamental layering of hardware and software tools predates Bratton’s philosophical treatment of it. The fundamental logic of the stack is that of layering and sedimentation of force relations, making it one of the most productive metaphors of the cybernetic imaginary. Much like Jeannette Wing’s theory about the layers which support computational thinking,^[10] the critique of computational thinking must also be layered, and must be modular along with the individual strata. It must in this sense dislodge itself from a hegemonic approach of outdated critical theory still largely embedded within a neo-Marxian paradigm in favor of a processual and generative network of relations. That means that critique towards computational thinking must become more nuanced and go beyond the luddite “Great Refusal,” and must rather adopt the complex and high-definition spectrum of discernment and know-how which is necessary to speak about computation and its potentials properly. It must know the language of power before it can venture to deconstruct it.

The Digital Humanities Stack

A valid critique of computational thinking can be found in the book Digital Humanities (2017) by David M. Berry and Anders Fagerjord in which the authors work towards a “critical digital humanities” which would offer “a pushback to the sometimes instrumental tendencies within digital humanities.”^[11] They go on to specify that part of this project “has to be the focus on the socio-technical aspects of technologies used, how they are assembled and made, and the possibility of making them otherwise.”^[12] The injection of the social into the largely abstracted and often hermetic discourse around digital technologies is what is at stake, and the digital humanities are poised to do just that. The effort is valiant, but unfortunately lags behind the cutting edge of corporate and military development which is often produced and enabled outside the realm of meso-political discourse. Apart from virile critique, it is thus necessary to understand the technological means and ends which may benefit society in parallel to the military-corporate war machine and to reclaim Deleuze’s belief that “technology is […] social before it is technical.”^[13]

Such a project needs to build stacks of its own, ones which may not be tethered to those infrastructures which are deemed hostile or dubious in their loyalty. Berry and Fagerjord offer their own stack: the “digital humanities stack” (DHS) which is built in order to facilitate the project of a “critical digital humanities.”^[14] This is one of many potential stacks to be built, but is interesting insofar as it offers a humanities critique of Wing’s computational thinking.

There are two segments in the DHS which I will focus on in the following paragraphs, and those are the lowest layer which couples “computational thinking” and “knowledge representation” and then the top-most layer where Berry and Fagerjord include the need for “critical / cultural critique” working along with “tools and apps,” “publications” and “projects” – all run-of-the-mill products of contemporary humanities academic departments. 

Meaning and Computation

The fundamental insight afforded by the coupling of computational thinking with knowledge representation in the bottom-most layer has to do with the relationship between information and meaning. 

Without getting into an overly detailed analysis, a fundamental primal scene for computational thinking were the years following World War II, when the field of cybernetics was becoming a central generator of metaphors which spanned far beyond the field of computation, and into biology, psychology, linguistics and philosophy.^[15] A fundamental insight was gained from the work of Claude Shannon and Warren Weaver, who formulated their highly influential “Mathematical Theory of Communication”^[16] which reworked the concept of information towards a binary exchange of standardized informational packets, or bits. However, in order for Shannon to abstract information into such discrete units, he had to strip it of ‘meaning.’ Margaret Mead and Heinz von Foerster were, among others, highly critical towards this understanding of ‘information’ as necessarily being stripped of ‘meaning’ already during the 7^th Macy conference in March of 1950.^[17]

Adopting an approach to computational thinking which is largely based in low-context, natural language, Berry and Fagerjord approach the computational turn from the direction of the humanities, which integrally work within the context of meaning. Meaning and the human observer are interlinked within their ontology, and as such constitute one of the two pillars of the digital humanities stack. Anthropocentrism in their take does not have to be read as a negative label, but is rather ambivalent in its implications: it can be understood as negative (as in the project of a vulgar form of posthumanism) or positive (predicated upon the assertion of the value of the human body and its habitus, as in the work of Wendy Lynn Lee and her evocation of the “anthropocentric optimum”).^[18] Berry and Fagerjord firmly stand on the side of meaning and consider it to structure and potentiate the ought of which automatons and digital prostheses potentiate. They in this way politicize technology and search for the generation of meaning which towards which it can employed.

The top-most layer of the DHS (the interface layer, or the layer where tangible output is produced) also accounts for the reflexive and critical approach towards the remaining three components. This critical approach is again largely connected to the meaning of productivity and within the context of digital humanities is intended to reflect on the practices which may be insinuated by the affordances of the digital mesh. As Alexander Galloway asks, “is it the role of [digital] humanities researchers to redesign their discipline so that it is symmetrical with [a digital] infrastructure?”^[19] and goes on to answer that “Few things will cripple the humanities more than the uncritical ‘adoption of tools’ or the continued encroachment of positivistic research methods borrowed from cognitive science, neuroscience, computer science, or elsewhere.”^[20]

Such an entrainment of thought complicit with Wing’s call to “automating abstractions” is represented as a moment when computation oversteps its bounds against the comparatively unmeasurable and largely black-boxed cognition of the human wetware. It is this humanist legacy of meaning that Berry and Fagerjord’s productivist project of digital humanities tries to retain through their call for the inclusion of ‘knowledge representation’ as counterweight to Wing’s computational thinking, as well as in their call for a critical approach to the affordances of digital technologies in the study (and defense) of the humanities. 

The coupling of the human to its originary technicity must in this way become productive through a decentralized a distributed construction of stacks which would only vicariously and partially index hegemonic megastructures. The strength of such a gesture lies in the interoperability of the stack as a stratigraphy of meaning, and in this way paves the way for what Rosi Braidotti calls the “post-humanities.”^[21] Such a project would aim to answer what becomes of the human sciences after understanding that it is indeed tethered to a dying animal who knows not what it is.

---

^[1] Jeannette M. Wing, “Computational Thinking and thinking About Computers,” Philosophical Transactions of the Royal Society (2008, No. 366) 3717. See also Jeannette M. Wing, “Computational Thinking Benefits Society,” Social Issues, accessed 1 January 2019 <http://socialissues.cs.toronto.edu/index.html%3Fp=279.html>.

^[2] Jeannette M. Wing, “Computational Thinking,” School of Computer Science, accessed 31 December 2019 <https://www.cs.cmu.edu/~15110-s13/Wing06-ct.pdf>.

^[3] Wing 2008, 3718.

^[4] See for example David M. Berry, “The Computational Turn: Thinking about the Digital Humanities,” Culture Machine (Vol. 12, 2011); David M. Berry and Anders Fagerjord, Digital Humanities: Knowledge and Critique in a Digital Age (Polity, 2017); James Bridle, The New Dark Age (Verso, 2019).

^[5] Wing was at the time Corporate Vice President at Microsoft Research, and had twice been Head of the Computer Science Department at Carnegie Mellon. She received her degrees from the Massachusetts Institute of Technology.

^[6] Louis Armand et al., “Perspectives on Computational Thinking and its Utility for Transformative Education,” Littreraria Pragensia (forthcoming).

^[7] Steve Easterbrook, “From Computational Thinking to Systems Thinking: A Conceptual Toolkit for Sustainability Computing,” in Proceedings of the 2^nd international Conference on Information and Communication Technologies for Sustainability (2014) unpaginated.

^[8] Easterbrook, unpaginated.

^[9] Benjamin Bratton, The Stack (MIT Press, 2015).

^[10] Wing, 2008.

^[11] Berry and Fagerjord, unpaginated.

^[12] Berry and Fagerjord, unpaginated.

^[13] Gilles Deleuze, Foucault (University of Minnesota Press, 2006) 40.

^[14] Berry and Fagerjord, unpaginated.

^[15] See James Gleick, The Information: A History, a Theory, a Flood (2011); or Tiqqun, The Cybernetic Hypothesis (2001).

^[16] Shannon and Weaver, A Mathematical Theory of Communication (University of Illinois Press, 1964)< https://monoskop.org/images/b/be/Shannon_Claude_E_Weaver_Warren_The_Mathematical_Theory_of_Communication_1963.pdf>.

^[17] Gleick

^[18] Wendy Lynn Lee, Eco-Nihilism: The Philosophical Geopolitics of the Climate Change Apocalypse (Lexington Books, 2016)

^[19] Alexander R. Galloway, “The Cybernetic Hypothesis,” differences (Vol. 25, No. 1, 2014) 126. 

^[20] Galloway 128.

^[21] Rosi Braidotti, The Posthuman (John Wiley and Sons, 2013).