Erving Goffman’s ‘paper’ — LessWrong