After E. de la Vaissière (Samarcande et Samarra, Leuven: Peeters, 2007: see my review infra in this section), the author rises up against the postulate according to which mamluks constituted the larger part of the armies of the Abbasid and of the Persianate dynasties of the Pre-Saljuq period. Through the study of chronicles, geographical works, and local histories, she shows that East Persian dynasties like the Saffarids, the Samanids, the Ghaznawids and others, as well as the Abbasids themselves, employed only a minority of slave troops in their armies. These slave soldiers were recruited for specific posts (as bodyguards and palace guards). If they were the cause of social troubles, they did not show militarily more efficient nor more loyal than free soldiers. Only when ruling dynasties were weakening, they tended to have recourse to massive incorporation of slave soldiers, for compensating their loss of military manpower. Slave soldiers subsequently contributed to the decline and fall of these dynasties.