Transliteration follows the guidelines of the International Journal for Middle Eastern Studies; therefore, only Arabic place names and terms not found in Merriam–Webster’s dictionary have been transliterated. In order to provide readers with a sense of the diversity of vocabulary used for different categories of slaves, the transliteration of Arabic terms for slaves was included in all my translations. Given the attention paid in this book to names and naming, all personal names were also transliterated.
Daniel Varisco has convincingly challenged the usefulness of the term “medieval” in reference to Islamic history.1 To accommodate readers not specialized in the Islamic context, the term will nevertheless be used in this book to denote the broad timeframe between 500 and 1500 CE. During this period, “Yemen” did not exist as a constant, unified political entity; the name is here used as shorthand for South Arabia, roughly corresponding to today’s Republic of Yemen, but not excluding neighbouring areas (e.g. parts of Oman and of south-western Saudi Arabia).
Throughout this book, dates referring to Islamic history are given according to both the Islamic (hijrī) and common-era (CE) convention, while dates of contemporary publications, scholarly debates and so on are only quoted in the latter.
For the purposes of this book, I translate al-Ḥabasha as “Ethiopia”, by which I mean the territories roughly corresponding to today’s Ethiopia and Eritrea. The term “Abyssinia” with which some scholars translate the term appears too narrowly identified with the Christian Abyssinian empire.
As per established convention, slave soldiers will be referred to as mamluks, while the eponymous Sultanate (648–922/1250–1517) will be capitalized (Mamluks).
“Making ‘Medieval’ Islam Meaningful,” Medieval Encounters 13, no. 3 (1 September 2007): 385–412. See also Thomas Bauer, Warum es kein islamisches Mittelalter gab: Das Erbe der Antike und der Orient (Munich: C. H. Beck, 2018).