[JoachimSchipper]

Nobody is arguing that intro-/extraversion does not exist and is not a valuable component in predicting human behaviour. pg 's article predicts that people self-identifying as X will have certain hangups around X, e.g. people self-identifying as shy are less likely to learn to be less shy, even if this would be desirable. Looking at this thread, that is exactly what happens. (And I'm done with this thread.)