This handsomely produced book is a study of one of the main genres of non-Classical Yemeni poetry, known locally as ḥumaynī. The origins of ḥumaynī are ancient, and it may well be a local adaptation of the medieval Andalusian and muwaššaḥ, types of strophic poetry which made their way to the eastern Arabic-speaking world sometime after the twelfth century ce. Formally, ḥumaynī is in stark contrast to the ancient Arabian tradition of non-Classical verse, known as nabaṭī, which has strict mono-rhyme and lacks division into groups of verses. A brief introduction is followed by the main body of the book, divided into five parts: (1) Chant Sanaani et Humayni; (2) La Langue du Humayni; (3) La Métrique du Humayni; (4) La Muʿārada; (5) Vers les Sources du Humayni. A conclusion and bibliography complete the study. A data annex of almost 200 pages follows (as is normally the case with French doctorats d'état), which lays out the Arabic texts of all the poems analysed in the body of the book, as well as a table of their metrical structure and a set of indexes to the poems and their authors.

Within Yemen, ḥumaynī seems to have developed in the medieval cities of Rasulid power situated on major trade routes to Egypt and Syria. Only from the sixteenth century, with the gradual shift of political and economic power to the north, did it begin to be practised in Sanaʿa, the focus of the present work. From its first appearance, it was associated with Sufism. And though there were never any constraints on its topoi, it was overwhelmingly, and even when composed by Sufis, a form of sung love poetry (ghazal) even if the “loved one” was God. It has retained this favoured topos to the present day. The corpus of no less than 540 poems, the basic data for the work under review (itemized individually in the annex), consists entirely of ghazal and covers more than four centuries, from the fifteenth to the late nineteenth century. This is a stunningly comprehensive exercise in formal literary analysis.

The language of the poetry, analysed in Part 2, is mixed, with dialectal elements rubbing shoulders with CLA ones. It contains dialect features from the Tihama, where it originated historically, which are no longer in common use in Sanaʿa (the future particle šā, the verb ḥād “to see”, the definite article am-). However, morphologically and lexically, it largely consists of the common dialectal patrimony of Arabic over a very large area, rather than the thousand-and-one peculiarities of Yemen's multifarious regional dialects. It thus bears the hallmarks of a literary koine formulated and “normalized” over many centuries. The author (rightly, in my view) rejects Mark Wagner's argument that any rhetorical intent (e.g. condescension via the use of dialectal Arabic) inheres in this mixing. “Mixing” is simply one of the essential features of its diction, as it is of other Arabic poetic koines of central and eastern Arabia.

Sometimes one could have wished for more regional contextualization. So, for example, the author's remarks on the use in his material of the important morphological feature of tanwin, which he correctly, but somewhat irrelevantly, notes is “a feature of Middle Arabic” could have benefitted from a wider reading of the now extensive dialectological literature on neighbouring areas of Arabia and its modern dialectal (so-called nabaṭī) poetry, which also abounds in this feature, and in exactly the same syntactic contexts as he notes for Yemen. In all the dialects (and unlike in CLA) it never marks case. In order to distinguish it from CLA tanwin, this suffix should be more accurately described as the adnominal linker of an indefinite noun and a modifying adjunct of some kind, most commonly a simple adjective. It occurs widely in central and eastern Arabia, but also in Sudan, in the Arabic dialects of central Asia, and (significantly for a poetic tradition which traces its ultimate origin to Spain) in Andalusian Arabic. It is common in the language of dialect poetry in all of these areas, even in regions where it has been lost in normal speech (such as Sinai). One also notices in the book linguistic parallels with southern Arabia which go unremarked by the author, such as the unusual Form V verbal noun patterns (p. 97). These have been described by Reinhardt for the “sedentary” dialects of neighbouring Oman, which have their ultimate historical origin in Yemen.

The subsequent sections of the book comprise a detailed treatment of the metrics of ḥumaynī poetry, a study of its historical origins, and a most interesting illustration of muʿāraḍa, a process by which the ḥumaynī poet would take the work of an earlier poet as a model and, by using the same rhyme scheme, metre, much of the same vocabulary and the leitmotif, weave a completely new poem based on it using these common elements, yet one clearly related to, and in a sense a tribute to, the earlier work. This process of creative imitation was widespread, and resulted in the creation of “families” or networks of poems sharing partially overlapping and common elements. Again, variations on this theme occur in the central Arabian tradition, in which poems written in response to others have the same metre, rhyme and much of the same vocabulary, though with the added twist that the words are sometimes used in a different (often contrary) sense to the way they were used in the original, or involve metathesis of the radicals. In both cases, Yemen and central Arabia, it seems that poetic creation involved shared tools and reworked elements, rather than the individualized artistic endeavour we are accustomed to in the West.

This book is a very fine piece of work and a most valuable contribution to the growing literature on Arabian dialectal poetry.