Key points are not available for this paper at this time.
I first discuss several strategies that serve as surrogates for theories in psychology: one-word explanation, redescription, drawing vague dichotomies, and data fitting. I then identify two conventions that allow these surrogates to flourish and multiply: null hypothesis testing, which makes precise hypotheses irrelevant, and the isolation of research in different disciplines, which prevents the exchange of positive metaphors between fields.
Gerd Gigerenzer (1998) studied this question.