Knox, J., Williamson, B. ,Bayne, S. (2020) Machine behaviourism: future visions of ‚learningfication‘ and ‚datafication‘ across humans and digital technologies.

Learnification : Learning as behaviour

With ‚learnification‘ Biesta criticizes the focus that the current economics of education places on the activities of students, as it completely ignores the context and the role of the teacher in the learning process. The teacher-student relationship is completely ignored, as if the teacher did not exist.
In the context of globalization and the upswing in educational technology, and especially through the use of machine learning and artificial intelligence, the focus is all the more concentrated on the student as a „customer“, as a manipulable subject in the course of the economization of education. So-called student-centred courses are being developed, which, however, do not refer to the learning activities of an autonomous, self-directing learner capable of rational thought, but rather to an emotional and social being whose behaviour is to be measured, controlled and influenced. What is measured through learning analytics is ‚activity and behaviour located in ’non-conscious‘ mental processes and autonomic ‚bodily‘ functions (Knox, et al., p. 35). Thus, the authors conclude that learnification and datafication of education leads to a return to behaviorism.

The platforms are designed to measure and student behaviour rather than reasoning or arguments. The shift happens towards the desire to detect and measure learning, ’not in the intentional expressions or decisions of a self-directing, auonomous learner

By analysing the student’s behaviour and through the corrective responses of complex machine learning technologies, it becomes clear that there must be a predetermined goal towards which the student should move. Thus, neither the interest nor the desire of the completely foreign student is taken into account, which could also be the basis for his or her learning goals. These are predetermined and not negotiable. Learning is completely geared to these given goals, there are no questions or fields of interest that are branching off, everyone follows the same path with the same goal. The learner – who should be the subject in a student-centered environment – in this consumerist view is considered the object of manipulation towards a preset goal.

(…) consideration of the educational software platform highlights important aspects of learning analytics that entail significantly different  ideas about learning and the learner than those assumed by a ’student-centered orientation‘: not only is data positioned before the desires of the learner as the authoritative source for educational action, but the role of the learner itself is also recast as the product of consumerist analytic technologies. (ibid.)

Behaviourism revisited

Reinforcement learning is the term which is emphasized by the authors to describe the process. The amchine takes over the (non-existent) role of the teacher re-naming it a ‚critic‘ who does ’not tell us what to do but only how well we have been doing in the past‘ (p. 37).  ‚This‘, the authors continue, ‚reveals a very different notion of learning to the autonomous neoliberal consumer defined by Biesta (2005), one that appears to strictly condition behaviour according to predefined values.‘ (37) In this way we come back to behaviorism and conditioning, the conditional training of certain behavior patterns. Right‘ and ‚wrong‘ do not refer to content (it is also debatable to what extent ‚right‘ and ‚wrong‘ per se are valuable feedbacks), but to a certain behaviour.

The authors then list several auxiliary sciences, such as behavioural psychology, neuroscience, behavioural economics etc), which have become ‚key vectors of governmental authority‘ in the last two decades and ‚human nature‘ has been reconceptualised: with Feitsma (2018a) they argue that the human (and therefore the student) is considered favouring immediate need and gratification over future planning, ‚with their decisions and behaviours involving emotional responses, habits, social norms, and the automatic, unconscious and involuntary  within human action, which might be nontheless predicted, regulated enhanced and exploited (38). Thus the learner becomes an ‚irrational and emotional subject whose behaviours and actions are understood to be both machine-readable by learning  algorithms and modifiable by digital hypernudge platforms.‘ (41) The student is thus incapacitated.By setting the learning goals and using the given skills training to learn a certain behaviour they become pure objects of behaviour technology. By setting the learning goals and using the given skills Training the skills to learn the desired behavior they become pure objects of behavioral technology.

Is there something like constructivist data science?

Learner centered models and concepts are not per se bad. But we have to include context, roles of student and teacher, other roles, learning objects, methods, self organized sequences and so on. The relationship between student and teacher is also very important when it comes to learning and education. We should probably to define (measurable) criteria what learning is. Constructivism does not exclude the teacher role from education, but demands a different role model from the traditional idea of the teacher. There are other roles: teachers who accompany and guide students on their (individual) learning paths, who arrange settings for the students to succeed, curating content and media in context to allow the students to learn, as we discussed with Selwyn’s texts.  During my education I was basically constantly told me what and how I had to do. If today learning analytics serve a behaviourist idea about education, what do we have to change?

If today only questions about effectiveness and efficiency of education are allowed we need to change this. Probably only a clear turning away from the neo-liberal concept, a change of the socio-political and economic base would be really effective. We must rethink our view of humanity and move away from eco-centrism and no longer allow neoliberal companies to determine what learning is.