Tajfel, H., Billig, M. G., Bundy, R. P. & Flament, C. (1971). European Journal of Social Psychology, 1(2), 149–178.
Not read. The summary is drawn from the abstract or from what the work is standardly cited for. A reading should check it first.
Why this is here
The founding experiment of social identity theory, which every fandom and social-identity reading in the record cites at second hand.
Read in its place
The work cannot be read here, so the record holds it at second hand.
Otten (2016) gives the design of Tajfel et al. 1971 at second hand: novel groups on a trivial criterion (art preference), anonymous allocation between other people on matrices built to separate parity, joint profit, in-group profit and maximum difference, and no way to pay oneself. It reports the result as a strong pull to fairness together with reliable in-group favouritism, and differentiation sometimes preferred to in-group gain. It reports the later critiques and replications. Missing still: the two experiments themselves, the participants, the matrices and their scoring, any numbers, and the authors’ own 1971 interpretation, which came before social identity theory.
Pinter and Greenwald (2011) give the classic procedure of Tajfel et al. 1971 only in outline: participants rate paintings (or estimate dots), are told falsely that this places them in a group, and allocate money to anonymous others, not themselves, with no interaction and no group history. Their classic condition is a modern version of that induction, and it again produced explicit favouritism, here on Bornstein’s matrices. They show that imagined or random assignment does as well on explicit measures. Missing still: anything of the original’s design in detail, its participants, its matrices, its numbers, and its own interpretation; the paper cites it but reports none of its results.