Awdaghost (the modern Tegdaoust) is the most extensively excavated medieval town of the Sahara, and, if we are to believe al-Bakri, was widely famed for its size and wealth. Located in south-central Mauritania, it was excavated between 1960 and 1976, with the principal results published in five successive monographs. Augustin Holl's interest in the site is of long standing, as he makes clear in his introduction, and fits within his earlier research in West Africa on the links between social dynamics and the spatial organization of settlements. The book's intention is straightforward: to take advantage of the detailed publication record for Awdaghost to analyse patterning in the use of domestic space across successive occupations of the site. As Holl points out, it is plausible that variation in how people organize their living space responds to demographic, social and economic influences. Its analysis should therefore open up windows on many facets of the past, including questions such as whether towns like Awdaghost were internally divided into ethnically or socioeconomically distinct neighbourhoods, how household units were structured and shaped, and how they evolved.

Chapter 1 places Awdaghost in its regional context and provides a useful review of work at other urban sites (though it is a pity that Hausaland is omitted from its scope). Chapter 2 then considers some of the theoretical background for developing an archaeology of households, which in the case of Awdaghost proceeds through the identification of three analytical units defined in the introduction: the household, the habitation complex, and the building sequence. A useful overview of the history of research at Awdaghost follows in Chapter 3, though the historical records for the site sit rather isolated from the rest of the text in Chapter 8. The bulk of the book lies in Chapters 4–7, which analyse the organization of space on the ‘Acropolis’ at Awdaghost from its still undated beginnings before ad 900, through a period of explosive growth and subsequent peak size and importance to its post-ad 1200 decline, a process that culminated with the site's abandonment about 1500. These chapters are, it must be confessed, nothing if not data-rich and provide an impressive level of empirical detail on the size, orientation and mutual connections of individual rooms, structures and spaces through each of the stratigraphic levels (‘Building Sequences’) into which the site could be divided. The text is amplified by an impressive series of plans illustrating the relationships described. Chapter 9, entitled ‘Patterns of Space Allocation’, then summarizes these data in terms of changes over time in the structure of public space (streets and squares), habitation complexes (sets of contiguous household units) and households (principally focused around variation in household size, courtyard size and room distribution). Key interpretative principles for reading these changes include the establishment of a permanent settlement at the very beginning of the site's history; the later switch from mud brick to stone as the main building material; the emplacement on the Acropolis of wealthier, larger households concerned with manifesting the Muslim concern for domestic (especially female) privacy; and the eventual impoverishment of the site that saw these complexes replaced by much smaller, more modest homes.

What we are left with, however, is nonetheless a largely descriptive account, and one constrained by its almost total dependence on formal readings of architectural organization. Strangely, given how much has already been published about Awdaghost, almost no reference is made to the artefacts found within the structures described. Hearths, in situ storage vessels, wells, latrines and a few benches may indeed exhaust the features present, and clearly assist in identifying function, but what of the rest of the material culture left by the site's inhabitants? Its incorporation would surely strengthen the ascription of room functions, which in many cases seem simply to be asserted. Moreover, one wonders whether more extensive and detailed reference could not have been made to wider patterns of domestic space organization in Islamic societies, as discussed by Tim Insoll and others, as well as to the practices of other Berber societies in northwestern Africa if, as seems likely, Awdaghost was occupied by a largely Berber population. The work of Giddens and others on the recursive relationships that exist between material culture (including architecture and space) and the ways in which people go about their daily lives and produce and reproduce social structures might also have been addressed. In short, this is a useful and detailed study, but one that falls short of its promise of an anthropologically informed household archaeology of a major West African town.