Minutes from May Meeting

Discussion of Barassi, and Pierlejewski:

• The main part of the discussion was on the Barassi chapter. It was considered a very good summary of datafication made into policy/put into practice. It shows a very realistic side of datafication, that is: people are often too tired from everyday life to resist datafication, if they care at all. The processes that datafication enables (and which enable datafication in turn) are simply convenient. People like their iPhones (as an example) and accept the asymmetries of capital that comes with this. However, Barassi did do a good job of showing where these conveniences fail and where datafication causes actual harm for people.
• Education is a slightly difficult (or strange) space to talk about such harm from datafication though. There has long been discussion of symbolic or benign violence in schooling (by Bourdieu, for example), as schools are where socialisation happens. So for datafication to facilitate techno-biopolitics in schooling seems logically consistent. Schools have always had a stratifying function so Big Data is only another step in this direction, rather than a quantified turn: the institutional processes of school cannot be disassociated with the processes of measuring. The question becomes, is education a field which operates distinctly from other fields in our society?
• There was then a discussion of datafication in contrast to manual measurement. In predigital moments, the combining of data points was done by a human. It may not have been carefully done but it was done by a human nonetheless. This suggests some amount of assessment of the data. In automation, any human anxiety around classification (and classifying) disappears.
• This automation also changes the data gaze. In the predigital, (educational) data was largely in an upward chain towards the state. Now everyone is both judge and accused; our responsibilisation for our data is both democratic and paranoid. You are tasked with always improving yourself, as well as always being watched. However, your data can still be black-boxed (or for parents, their child’s data can be black-boxed) so we have not achieved some totally transparent system.
• The question is if datafication has a tendency towards asymmetry and monolithic power or if it can be shared and held in common? This is a fundamental question about the deconstruction of hegemonic power and relates to Raymond Williams’ analysis of broadcasting. This also brings us back to the importance of alternative perspectives on data, such as for Indigenous groups or economically poorer groups. Can data be used to show the validity of their claims about society or is data structured in such a way to ensure the status quo? If any progressive points are accepted, do they then get cannibalised by hegemonic discourses, rather than bringing about structural change? 
• There was then a discussion of alternative systems (or possible alternative systems). If we (hypothetically) had an education system with a sample size of one student (N=1), how would we measure this student? Or would we at all? Or if we had a system in which people were trained to replace particular people, would comparison between students become redundant or impossible? It was felt that these systems would still stratify by other assessments (such as class or caste) and that rigid, functionalist societies ultimately implode. Datafication works because it only exercises a soft power over people (even if this soft power is very influential).