Data-driven schools – the end of teacher subjectivity?

Early on in his new book on automated media, Mark Andrejevic (2020) offers an interesting perspective on the promise of data-driven automation … and the corresponding challenge it poses to teachers. As Andrejevic puts it, data-driven automation introduces a new form of machine-driven subjectivity into our lives that circumvents what are perceived to be the inevitable failings of human decision-making … i.e.

“the moment of uncertainty, unpredictability, inconsistency, or resistance posed by the figure of the subject. The problem stems from the fact that subjects can be unpredictable, recalcriant, and otherwise irrational in ways that threaten systems of control, management, and governance. An automated subject would allow a fully automated society to run smoothly and frictionlessly – whereas actual subjects threaten to gum up the works” (Andrejevic 2020, p.2)


New digital systems are therefore predicated around an imagined ‘version of AI’ that is all-powerful – intelligent technologies “that can make sense of the world in ways that humans themselves cannot”. The combination of powerful computational techniques and massive data-sets therefore promise to transcend the limits and inefficiencies of human information processing.

This urge to automate comes from the integral roles that human subjects play in areas such as consumption, production, politics and security – all of which have come to be fundamental to social, economic and political underpinnings of society. Thus the complex and unpredictable nature of humans is seen to introduce frictions and resistances that slow social and economic processes. In contrast, digital automation promises to transcend “the internal tensions that accompany the divisions in the subject: between conscious and unconscious, individual and collective, culture and nature” (p.4). As Andrejevic puts it, “finally, all can be known: no more doubts about human or natural risks” (p.2).

Andrejevic also raises the interesting contention that the appeal of digital automation stems from the idea that these technologies do not simply offer faster thinking and decision-making, but thinking that is fundamentally clearer, more pure and objective. As such, this is seen to result “in the abstraction of a task away from the motivations and intentions in which it is embedded” (p.4).

Of course, it is important to look beyond this ‘fantasy’ of ‘total information capture’ and pretentions of knowing ‘everything’ about the world (p.20). Alongside the practical limitations of digital technologies and their capacity to capture data, Andrejevic also raises the fundamental pointlessness of redoubling the world in data. As with Borges’ story of Funes the Memorious – the man who could not forget – knowing or remembering everything about the world would be of no practical use. As Borges put it,  ‘To think is to forget differences, generalise, make abstractions’ (1944 p.154). In contrast, with his complete knowledge of the world, Funes becomes paralysed. Extending this logic to the present day, what could a datafied complete facsimile of the world tell us that we could not already know through first-hand experience? The concern is that too much data will actually limit our capacity to interpret and abstract meaningful information from it. As Andrejevic argues, we are ‘creatures of narrative’ – meaning the ‘holes in our memories make possible our accounts of ourselves, and our experiences of the world’ (p.34).

These observations certainly hold true in the case of teachers and schools. Education is fundamentally based around human relations (which are mostly unquantifiable). Education also requires teachers and adults to forget and let go of expectations and assumptions about students in order to foster development. A datafied school (and datafied subjects) can only ever present a very small slice of the essentially human and personal nature of schools and schooling. Clearly there is a role for data in informing what goes on in classrooms, but the drive to erase human/teacher judgment (and, by extension, the human/teacher subject) in datafied processes is ultimately counterproductive.

That said, it is important to remember any digital automation is not wholly abstracted and decontextualized. Indeed, datafication is most likely to result in a reconfiguration of the task (and the humans involved in it) that fits the motivations and intentions of a system’s designers, programmers and developers. In this way, automated systems result in narrow definitions of human subjectivity – thereby running the risk of ‘diminishing our subjectivity in order to comply with digital systems’ (Andrejevic, 2020, p.133).  All told, it is understandable why many teachers might find the promise of data-driven automation in the classroom as fundamentally unappealing.



Andrejevic, M. (2020).   Automated media.   London, Routledge.

Borges, J. L. (1944/1964). Funes the Memorious (J. E. Irby, Trans.). In Labyrinth (2nd ed.). New York: New Directions.