Arabs and Empires is a compendium of inscriptions, archaeological finds and Late Antique literature about pre-Islamic Arabia, making it an expansion and update of the field's previous sourcebook, Hoyland's 2001 Arabia and the Arabs. Fisher mobilized a range of experts, and the book contains excellent contributions on Arabian epigraphy by Michael Macdonald (11–56), South Arabian inscriptions by Christian Robin (90–171), Ghassanid/Jafnid archaeology by Denis Genequand (173–207), and a handy chapter by today's leading scholarly voices on the array of Arabian languages from which Old Arabic emerged (373–433). These empirical sections constitute the book's principal strength: they cover essentially all salient archaeological discoveries to date, and students will find much for their imaginations and interest. Sections analysing literary sources and the overall management of the material, however, tread uneasily over sensitive theoretical matters of social identity, Arabness and ethnogenesis, and thereby produce a somewhat problematic vision of Arab history raising a number of issues.

To begin from the literary selections, non-Arabic Late Antique records are expressly privileged over Arabic, on the premise that Arabic accounts were only written after the fact – in the Muslim era. This makes Arabs and Empires something of a full pendulum-swing from Jawād ʿAlī, Irfan Shahid and Aziz al-Azmeh, whose histories place substantial reliance upon Arabic literature. Whilst Shahid and al-Azmeh have been criticized for uncritical employment of Arabic reports, prudence would seek a middle position. Although Muslim-era writers indeed reshaped historical memories, the extent of their reconstructions does not sustain Fisher's surprisingly bald assertion that Muslim writers were “outsiders” (p. 2), prompting his editorial decision to relegate Arabic literature to the last chapter as a restricted addendum. Methodologically, the conceptual category of “outsider” is questionable because not all putative outsiders are the same: ancient Greek commentators had limited contact with Arabia, Syriac monks apprehended Arabian Saracens as an uncertain barbarian menace in need of conversion, whereas early Muslims were describing their own community's history, and as such, they negotiated memories which had an organic genesis from earlier Arab communities, and which retained unique relevance for the lore's recorders.

Whilst acknowledging the hyperbole and topoi in hagiographic writing about Saracens, Arabs and Empires operates with a strong current of acceptance of the greater empirical truth in Late Antique literature than in Muslim-era Arabic. Aside from the significant point that it is obtuse to tell the story of the Arabs without recourse to Arab voices, the oft-repeated reference to Muslim-era writing as a single, homogenous “Arabic tradition” is concerning. Arabists have long remonstrated that this is the incorrect way to approach Arabic literature since the corpus is refreshingly heterogeneous. Pre-Islamic poetry describes the varied contours of tribal communities, belief structures and the broader people of Maʿadd; the layers of Arabic literature reveal debates over Ishmaelite genealogy and the complex path by which Yemenis became “Arabs”; and careful studies of pre-Islamic prosopography have been fruitful. Fisher's decision to eschew Arabic testimony overlooks key insight into the nature of Arab community and identity.

Byzantine and Syriac writings are very valuable, and Arabs and Empires recounts an expansive selection, but the skewed emphasis entails manifest risk that non-specialist readers will conclude that Arabic accounts about Arabs are of limited worth. Critical reading of the Arabic Jāhiliyya material is actually a pressing desiderata which introductory surveys ought to encourage – the presentation of Arabic materials at the end of Arabs and Empires, far from where its argument lies, unfortunately engenders the opposite impression of the sources' utility.

Chapter 6 on Christianization is revealing of the book's value and limitations. On one hand, it is a compendious selection of carefully analysed hagiographies, but it tends to a one-sided view that affords scant room for contemplating the broader picture. Its thesis about the politics of sectarian division is currently contested, and in the interests of crafting a balanced narrative, Tannous' contributions should have been voiced. Furthermore, Cheikho's Arabic primary material is absent, so readers perceive Christianity's spread through the lens of hagiography. Greek and Syriac texts are problematized in numerous respects, but since other voices are scarcely heard (the Ghassanid inscriptions included are intriguing but laconic), readers are left with Byzantine worldviews, thereby perpetuating faulty impressions about pre-Islamic Arabian faith. In this vein, we read that St Sergius was a key saint for “Arab Christianity”, since this is what outsider monks optimistically believed, but a survey of Arabic literature provides important rebuttals, e.g. the name Sarjis (widespread in sixth-century Greek and Syriac) made no inroads into Arabian onomastics.

Monastic writers may have known the fringes of the Syrian Desert, but extrapolating them to interpret a pan-Arabian Arab people is perhaps Arabs and Empires' primary difficulty. Greek/Syriac sources speak about Saracens and Ṭayyāyē, never “Arabs”, yet Arabs and Empires assertively assumes all are references to “Arab”, “peripheral people” – whatever their names in the sources – inviting readers to presume that pre-Islamic Arabia was full of somewhat homogeneous Arabs. Stroumsa's thesis (“People and identities in Nessana”, 2008) deconstructed Palestinian Byzantine writings, revealing that their use of “Saracen” did not connote one ethnic group, let alone one ‘Arab’ people, but Arabs and Empires seeks to argue that Arabs were a ‘barbarian’ ethnos expressly likened to the Goths or Franks, and so omits Stroumsa, as well as Fergus Millar and Fred Donner's recent critiques of generalizing pre-Islamic Arabians as “Arabs”. Again, an introductory sourcebook benefits from highlighting scholarly disagreement to prompt critical investigation, but Arabs and Empires overlooks theoretical conceptions of ethnicity in favour of whitewashing pre-Islamic Arabia with “barbarian Arabs”. Brief provisos (4–5, 89) note the “Arab” ethnonym's complexities, but these are forgotten in the insistent presentation of “Arabs” across the narrative chapters 5 and 6. Such provisos should have had what Lord Denning called a ‘red-hand’ alerting readers to the limitations and ramifications of the editorial decision (Spurling v Bradshaw [1956] 1 WLR 461). Arabs and Empires is thus a worthy sourcebook on archaeology and Arabia in the Byzantine imagination, but in leaving Arabness and Arabic literature in the margins, it is a book about Empire, not Arabs.