‘Machines don’t have instincts’: Articulating the computational in journalism

Abstract

This article examines the articulation of computational journalism, focusing on how the meaning of the computational is discursively constructed and mobilized as a specific constellation of intelligibility within news organizations. Relying on the concept of articulation developed in cultural studies, the article asks what, exactly, is meant by the computational in the context of journalism? Drawing on interviews with key managerial staff, editors and developers at Scandinavian news organizations, three broad claims about the linkage between the computational and journalism emerged. These articulations include the notion that machines don’t have instincts, that democracy can never be personalized and finally that the computational is something to think with, rather than simulate. The argument is made that what can and cannot be calculated is not merely a technical question, it is also a deeply social, cultural, political and economic one. Thus, the computational emerges as an important organising framework and discursive order for thinking and talking about journalism in the digital age.

Hybrids, materiality, and black boxes: Concepts of actor‐network theory in data journalism research

“…Relations between humans and nonhumans are informed by processes of association and translation that can be “material as well as social, physical as well as semiotic” (Michael, , p. 41). The sociomateriality of technology—a concept increasingly mentioned in journalism studies (Borges‐Rey, , ; Bucher, )—comprises scripts

“or rules of use: in order to get them to ‘work', particular capacities and skills need to be in place. In other words, human comportment needs to adapt to the demands of the technology” (Michael, , p. 41).

…”

“…At the same time, and in contrast to theorisations of data journalists' hybrid nature, interviewees' articulated “a subtle aversion to adopting computational thinking” (Borges‐Rey, , p. 14) while acknowledging its importance for data‐driven newswork. Another approach is presented by Bucher () who explores computational thinking through the cultural perspective of discourse. The author observes “that algorithms and software are entangled in practice.…”

“…This is to say that the influence is not one‐directional. Rather, journalism takes place in relation to, and within, material arrangements and discursive formations” (Bucher, , p. 929). Bucher presents accounts of journalists that acknowledge the role of algorithms and software, not so much as mediators of computational skills, but of computational thinking: Journalists had “developed a ‘feeling for the algorithm'” (Bucher, , p. 929) and recognise that their endeavours had not been achievable without computation.…”

“…Rather, journalism takes place in relation to, and within, material arrangements and discursive formations” (Bucher, , p. 929). Bucher presents accounts of journalists that acknowledge the role of algorithms and software, not so much as mediators of computational skills, but of computational thinking: Journalists had “developed a ‘feeling for the algorithm'” (Bucher, , p. 929) and recognise that their endeavours had not been achievable without computation. At the same time, discursive strategies were used to marginalise the agency of computation in order to emphasise the need of (human) journalistic expertise (Bucher, , p. 931; see also van Dalen, ; Young & Hermida, ).…”

“…Second, technological actants are to be examined whether or not they have an effect on an actor's newswork and how they are embedded within "the entire organizational assemblage of journalists, businesspeople, and technologists" (Lewis & Westlund, 2015, p. 24). Several studies have explored how technological actants affect internal dimensions, the interplay of data tools and practices and epistemologies (De Maeyer, Libert, Domingo, Heinderyckx, & Le Cam, 2015;Fink & Anderson, 2015;Karlsen & Stavelin, 2014;Parasie, 2015;Parasie & Dagiral, 2013); external dimensions via material implications of the outcomes of data-driven newswork, such as visualisations (Appelgren, 2018;Lowrey & Hou, 2018) and a combination of internal/external dimensions (Borges-Rey, 2016;Bucher, 2017).…”