The paper is concerned with the limits of psychoanalysis and the concept of the non-analytic, in the light of the work of Lacan and Derrida and the relation to Freud. The functions of rumour, gossip, third parties are discussed, so as to highlight structural features of the analytic situation; such that the analytic scene is always the other scene. The function of repetition that this necessarily supposes is examined, in the light of Derrida's theory of the name of Freud, of the filiation to Freud, and the historical origins of analysis. The central concept of transference (and its recent twin, countertransference) are linked to this repetition of the other scene, and Lacan's attempts to propound a theory of stories about becoming an analyst is shown to be one reponse to the impasse of the unanalysed residue left, and represented by, Freud.