This article presents the results of an exploratory research on how
ordinary people talk about algorithms publicly, and in so doing perform aspects of
their identities. To do this, we look at messages posted on Twitter in 2017
containing the terms 'Facebook algorithm'. From a qualitative content analysis, we
identify three basic types of " discursive algorithmic characters", that is to say, the
subjective positions in which the person decides to act when talking about “the
algorithm". They are: the critical subjects, the represented subjects, and the agent
subjects. We contribute to the current literature by showing how ordinary people
construct discursive identities in relation to the “algorithm”. In the end, we raise
three hypotheses to be further investigated: algorithmic identities may be
consciously co-constructed, algorithms are consumable cultural products, and
algorithms structure new types of audiences.
