This book was dealing with a highly polemical issue more than a century after the reference study on the Mozarabs was published. It was an attempt at combining textual with archaeological data, Arabic with Latin sources, Andalusi with North Iberian texts, edited sources with the unpublished corpus of Arabic glosses in Latin manuscripts. The published reviews bear witness of how the book was wellcomed : Les Annales (2012/1), 202-4 ; Speculum 87-1 (2012), 177-9 ; Hispania vol. 70, nº 236 (2012), pp. 797-802 ; Edad Media 13 (2012), 301-3 ; al-Qanṭara 32-2 (2011), 569-73 ; Medioevo Latino 33 (2011), 1026-7 ; Collectanea Christiana Orientalia 8 (2011), 313-22 ; Hispania Sacra 127 (2011), 788-90 ; Anuario de Estudios Medievales 41 (2011), 901-2 ; Le Moyen Âge 117 (2011/1), 192-3. Since Simonet turned the Mozarabs into the symbol of a Hispanic identity forged in the resistance to Islam, Spanish arabists, in response, have tended to marginalise that minority in the history of al-Andalus, going so far as to assert that from the 9th century on the Mozarabs survived thanks to the infusion of bishops and settlers from the north. To decipher this controversy one must go back to the Latin writings of the 12th-13th centuries, which attempted for the first time to identify this Arabized Christianity which, having been expelled from the Islamic lands, by then survived only in Toledo and a few places of exile. The rarity and dispersal of the existing sources have prompted some authors to place excessive reliance on argumentum a silentio. The drying-up of Latin writings produced in al-Andalus after the 860s lent some credence to the idea that the movement of the Cordoban martyrs was the swansong of Mozarabism. In fact the reverse is true: it marks the take-off point, for it was in the second half of the century, at Cordoba, that an Arab-Christian culture began to take shape. The object of this study is to throw light on a period which, although obscure, is the very crucible of the Mozarab phenomenon, defined as the gradient of Arabization of the Christian population of al-Andalus. The first part of the book shows that this inflexion was a reflection of the deep-going Islamicization of society, the first sign of which is the dismantling of ecclesiastical structures, which gained momentum in the second half of the 9th century. However the dismemberment was uneven, for a fairly cohesive ecclesiastical network survived in Betis until the 12th century. In the Gharb al-Andalus and the Ebro valley, where evi-dence of religious institutions is rare, some communities nonetheless survived until the rise of the Almoravids, who conducted the first expulsions. The advent of " neo-Mozarabs " may have revived certain groups, but the reinsertion of populations and institutions under the aegis of the North is a product of fiction. On the other hand, there was some remodelling within the body of southern Christendom. The 9th cen¬tury marks a turning-point inasmuch as Christians were then reduced to a minority by the transition to Islam through conversion and marriage alliances. The writings on the Cordoban martyrs, the Caliphate's discourse on fitna and the first Muslim legal sources together portray a society in flux, where conversion was not yet accompanied by social and cultural reorientation at an individual level, giving rise to the appearance of these half-way groups, the Muwallads or the " occult Christians " of Eulogius of Cordoba. At that time Christian apologetics, like the discourse of the Malekite fiqh, made significant efforts to define and codify inter-community boundaries. Of the distinguishing marks of Christianity, language took pride of place, becoming the subject of debate between the school of thought that was intent on saving Latin and the advocates of turning to Arabic as a new evangelising vector. Latin remained a cul¬tural emblem at the height of Arabization, as evidenced by its survival as the language of liturgy and the church and its continued use in funeral inscriptions. The advocates of Arabization endeavoured, while refuting the i'jāz dogma, to demonstrate that the Church should adapt to the linguistic practices of its worshippers. And thus a translatio studii - contemporary with Eastern Arab-Christian literature and not without influence in the Peninsula - displaced the literature of martyrs. Composed chiefly of biblical and religious translations, it did make some original contributions, traces of which survive, especially in the field of theological confrontation with Islam. As for the unedited body of notes in Arabic scattered through a good number of Latin manuscripts, their analysis simply highlights the depth of Arabization. However, despite the promising beginnings and despite persisting through to the time of the Caliphate, this movement played only a marginal role in the forging of Andalusian Arab culture. This was surely a consequence of the erosion of Christianity, but also of the incompatibility between the Umayyad model, locked in rivalry with the Arab East, and an Arab-Christian culture still tethered to the pre-Islamic model. Mozarabic culture, then, shared the same paradigm with the northern Iberian societies. That explains the ease with which migrants from Hispania were integrated, encouraged by the local powers to take part in the restructuring of the marches reconquered from Islam. These movements are evidenced by narratives and inscriptions, by the circulation of books, by the use of Arabized anthroponyms, and even by instances of the use of Arabic in clerical annotations. This influx is undoubtedly not the generalised exodus depicted by Astur-Leonese sources to project the image of a monarchy of resto¬ration. There are a number of further complex factors that explain the proportion of Roman-Arabic anthroponyms in the documentation, which reached an average of 30% in the Beira region by the 12th century. Migrations spread onomastic usages previously acquired in al-Andalus, but their density in frontier areas was also the result of con¬tinuous contact with Islamic territories. To the residues left by the first phase of Islamic occupation were added a welter of relations with the South, which exerted a veritable cultural magnetism until the 11th century. This Mozarabic situation, characterised by the selective integration in Latin cultures of elements borrowed from the symbolic rep¬ertoire of Islam, disappeared in the 12th century.