On first reading, Leiris's posthumously published Journal 1922–1989 resembles a massive collection of ébauches that helped produce the eventual autobiographies which we have already read. This circuitous relation tends to create a combined sense of confirmation and destabilization, a lulling déjà lu unsettled by a vague Unheimlichkeit. This occurs above all when we read the record of a physical or linguistic event that is to be elaborated later in a work which we have in fact already encountered; the chronology of our ‘Leiris’ is suddenly contradicted by Leiris's most intimate voice. Thus, for example, we encounter the 1955 Fourbis's Khadidja at the physical rather than textual moment of her meeting, in 1940 (J326); or we find that the ‘Dernières paroles’ section of the valedictory Frêle Bruit of 1976 (FB290) was in fact written in 1970 (J645–6). One's first reaction, I feel, is unconsciously to neutralize this potentially disruptive effect. Rather than opt for estrangement, so allowing the Journal to overturn ‘Leiris’ and become the matrix of a somehow new and original reading, the instinct is surely to absorb the journal's entries into the framework of the established ‘Leiris’, assessing them consequently as the raw material for a later but already familiar performance. The journal itself encourages this view in its self-characterization as formless and unfinished.