Jonathan Brown's doctoral dissertation (University of Chicago, 2006), now his first book, describes how the collections of Bukhārī (d. 256/870) and Muslim (d. 261/875) acquired such prestige as to constitute a virtual canon among Sunni Muslims. In Part One, he describes Bukhārī and Muslim themselves, reviews the field of ‘canon studies’, and describes how Bukhārī's and Muslim's collections became canonical by the beginning of the eleventh century. The description of Bukhārī's independent, distinctly anti-Ḥanafi juridical stance is notably effective. It transpires that the scholars who studied, replicated, and exalted Bukhārī's and Muslim's collections above all others tended to be Shāfi'i in law and roughly Ash'ari (i.e. middle-of-the-road) in theology. The outstanding example and single most important contributor to the process of canonisation was evidently al-Ḥākim al-Naysābūrī (d. 405/1014). Thus two collections of hadith on which all Sunnis could agree emerged concurrently with four mutually recognised Sunni schools of law. Agreement over Bukhārī and Muslim to some extent mitigated disagreement over the law.

In Part Two, Brown describes how Muslims have treated Bukhārī and Muslim from the High Middle Ages until the present, justifying or questioning their canonical status, exploiting or ignoring them in the course of juridical controversy. Ḥanafi jurisprudents were notably slow to recognise the new canon. Jurisprudents of all schools tended to use them opportunistically, proudly describing a hadith report in their favour as muttafaq 'alayh if both Bukhārī and Muslim included it but freely working around such a report, not modifying their school's traditional position, if it apparently went against them. A particularly interesting chapter recounts how iconoclastic Muslims, notably Nāṣir al-Dīn al-Albānī, have recently raised new doubts about Bukhārī and Muslim, in response to which indignant conservatives have asserted their impeccability more emphatically than ever.

Brown disavows any intention to test the authenticity of hadith, but of course the issue does arise. He reports correctly that the most widely accepted distinction between Bukhārī and Muslim is that the former accepted links only if the transmitter and his shaykh were positively known to have met, whereas the latter also accepted links if transmitter and shaykh definitely might have met. Brown does not himself investigate how he knew who had met whom. He does point out some hadith experts who observed that whereas it might be known that two persons had met once, it could not be known that everything one related from the other had come from that session, also experts who expressly called for not looking into the point too closely, lest it be found to have no basis. Brown implicitly sides at the end with the skeptics.

In a first book covering so many centuries, it is probably inevitable that some peripheral issues are not treated adequately. Appeals to ijmā on top of tawātur in defence of hadith certainly deserve study, but Brown cannot be said to do much more than raise the issue. Similarly, like Hallaq before him, Brown notices a discrepancy between the theories of hadith specialists and of jurisprudents writing about hadith but does not develop it in detail. Dates are often carelessly converted from Hijri to CE, too many non-standard, commercial editions are cited in the notes, and translations are often strange; e.g. ‘people with agendas’ for ahl al-ahw?'.

Brown's location of Bukhārī on the juridico-theological spectrum of his time seems to me particularly unsatisfactory. At one extreme were what have usually been called ‘traditionalists’ (Brown refers to ‘transmission-based scholars’): those who looked mainly to hadith for answers and expected hadith reports to speak for themselves. At the other extreme were ‘rationalists’ (Brown: ‘reason-based school’) who gave considerably more weight to reason in making out God's will (also to the Qur'an). Like the leading tenth-century enthusiasts for his work, Bukhārī adhered to a middle party that used rational means to compare and rank Qur'anic passages and hadith reports in law and to defend traditionalist tenets in theology. For conceding that one's pronunciation of the Qur'an was created, not co-eternal with God, Bukhārī was expelled from Nishapur by the traditionalist leader there, Muḥammad ibn Yaḥyá al-Dhuhlī (d. 258/872), later also, by one report, from his native Bukhara by local traditionalists. Brown argues at length that Dhuhlī's hostility came of personal jealousy, but this can hardly explain how he persuaded so many others to reject Bukhārī. Nor does it explain why Bukhārī composed a theological book (Khalq afāl al- 'ibād, ‘the creation of human actions’) to defend himself rather than simply pointing out Dhuhlī's conceitedness. Regrettably, Brown overlooks the report of Bukhārī's being expelled from Bukhara for theological reasons, mentioning only the alternative story that the governor expelled him for not teaching his children in a special session. He likewise ignores the report that the leading traditionists of Rayy dropped Bukhārī s hadith on account of his theological position.

Brown will not take these arch-traditionalists seriously, apparently from not wanting anybody to outdo Bukhārī in loyalty to hadith. He invents a new term for them: ‘über-Sunnis’. Hodgson objected to ‘traditionalist’ on the ground that their programme was sometimes demonstrably innovative. More than one American journal actually forbids authors to use it, although they have not proposed any alternative term. I cannot say whether journal editors will accept ‘über-Sunni’, but I will say it sounds irresponsible to use an obviously pejorative expression. Brown is willing to talk at the end of Madhhab Traditionalists and Traditionalist Salafīs, so ‘traditionalist’ evidently suits him when it does not threaten Bukhārī 's credentials.

The golden age of hadith science was obviously the time of the great compilers. In its silver age (11th–15th centuries), hadith was the most popular of the Islamic sciences, at least as measured by number of participants, but it has attracted far less scholarly attention than the golden age. Brown's survey establishes some important reference points and so advances the field.