The western study of ḥadīth has rarely viewed the ḥadīth collection in literary terms. Some specific ḥadīth have been treated as literary texts in their own right (the ‘ḥadīth of Heraclius’ is a notable example); but the literary merits of large collections of ḥadīth have not often been disucssed. Those ḥadīth collections that have been explored as ‘literary’ texts have tended to be those that are firmly part of the medieval Arabic literary genres of belles-lettres (adab) or history (taʾrīkh); works such as the collections of Ibn Abī ‘l-Dunyā, Ibn Qutayba, Ibn ʿAbd Rabbīh and Ibn ʿAsākir. Stefan Leder has dedicated a number of studies to the use of ḥadīth and akhbār in historical and literary works, and the way in which ḥadīth/akhbār are used in these contexts. Leder illustrates that elements of ḥadīth/akhbār are used in subtle and sophisticated ways. In contrast to scholars such as Leder, ḥadīth collections are often regarded as repositories of information, rather than as texts in and of themselves. The aim of this article is to explain why it can be helpful to approach all ḥadīth collections as literary works, rather than as treasure stores of information that can be raided as and when necessary for supporting evidence concerning this or that idea. This is particularly prominent in any studies that use ḥadīth to explore the idea or view of Islam on a certain subject (e.g. ‘martydom’, ‘eschatology’, ‘Adam’ etc.). Such studies have an important place in scholarship, but, I argue, it is also important to reflect on ḥadīth collections as a whole, as well as the meanings of individual ḥadīth within the wider frame of a collection.

This article will argue that ḥadīth collections generate a discourse through the selection and arrangement of the ḥadīth that are contained in an individual compilation; and that meaning is not simply given through the statements included in the individual ḥadīth themselves. The creation of a ḥadīth compilation is the bringing together of fragments of a larger whole (the ‘ḥadīth literature’ or ‘ḥadīth corpus’), into a new compilation. The idea that ḥadīth compilations are concerned with the creation of a specific discourse is an idea that has been put forward before, particularly by Louis Pouzet and Andrew J. Newman,7 but it is one that neither Newman nor Pouzet have developed in great detail, and it is certainly an approach to reading ḥadīth that has not gained much attention in Ḥadīth Studies more widely.