Canterbury Cathedral's Trinity Chapel, as rebuilt after the fire of 1174 to house the shrine of St. Thomas, was one of the most cosmopolitan and influential achievements of the Gothic age. This study considers not the architecture, but the pavements and fittings, of the east end of Canterbury Cathedral, particularly the geometric mosaic Opus Alexandrinum pavement before the site of the shrine, its related pavements, the shrine base, and the adornment of the corona at the cathedral's east end. The authors, having justified the need for a study of this material, begin with a close archaeological description. They argue that the mosaic pavement was retained from the pre-fire church, that its form is Romanesque not Gothic, and that it may have been set down as late as the period 1230–1250, long after the translation of St. Thomas to the shrine in 1220. They therefore maintain that the pavement's relocation falls closer in date than hitherto thought to the Cosmati pavements and monuments set down in Westminster Abbey (also recently studied by the authors), beginning in the 1260s, and that the two projects were intimately linked. These conclusions are radical, and to many will be surprising.

The strength of this book lies in the beauty of its production and in the thoroughness of the descriptions it offers of the literature and of the works themselves. It would be hard to imagine a more comprehensive illustrated account of what is physically in the Trinity Chapel. However, the central weaknesses of this book are analytical. Conscious limitation of scope is one issue: we find nothing on politics or liturgy. The fact that mosaic work of this type was common in Byzantium, the Holy Roman Empire, Italy, and France is not sufficiently emphasized. In form and material, the pavement is regarded as sui generis (35), whereas its genre was in fact widely established. Other propositions can neither be proved nor disproved. For instance, there is no evidence for the authors’ idea that this particular pavement survived in its present form from the pre-1174 church. In a passage of ekphrasis on the old church cited in the book, William of Malmesbury alludes to its “shining” pavements, a rhetorical commonplace, not archaeological evidence. The authors may be right—if unoriginal—in suggesting that the new building reused mosaic work from the previous one. Yet their notion that an entire pavement was preserved and then relaid in the new building is pressed forward unrelentingly. In defining (questionably) the pavement's design as “Romanesque” in style (87), the authors suggest that the biggest repertory of geometrical designs of this type at Canterbury, the iron frames in the chapel's stained-glass windows, derived their ideas from the floor and not vice versa. This cannot be correct. The geometry of four large circles bisected by the sides of the central square of the pavement is also found in the related Gothic windows at Sens Cathedral, but these will not have followed a floor at Canterbury. The generative importance of the ferramenta is also apparent in the design of the c. 1200 primatial marble throne at Canterbury. Omitting a full discussion of the problem, the authors present an excitingly radical but implausibly complicated alternative based on parti pris logic.

Throughout this book we constantly encounter the same baffling mix of admirable inventorizing but tendentious discussion. This becomes more problematic as the book widens out to embrace the Roman mosaics at Westminster Abbey, begun late in the reign of Henry III. Many commentators have agreed that Canterbury's standing will have influenced shrine churches such as St. Edward the Confessor's at Westminster, not least because Henry III attended the translation of Becket in 1220. The authors, however, have an agenda, which is to close the gap between the very different mosaic works at Canterbury and Westminster (so-called Cosmati works signed by Odoricus and his son). To do this, the reinstallation of the Trinity Chapel pavement is postponed to the period 1230–1250 (317), within a decade or two of the inception of the Westminster mosaics in 1266–1267. The (inexplicably) delayed re-siting of the Trinity pavement in the period 1230–1250, long after the 1220 translation, is stated by the authors to have provided the “springboard” for Westminster (338). Because the Westminster Cosmati masons must have received training in using Purbeck prior to their employment at Westminster, this could only have occurred at Canterbury (340).

Little of this inspires confidence. There are serious reasons for believing that the stimulus for the Cosmati mosaics came not from Henry III, but from Pope Clement IV and the papal legate Ottobuono Fieschi, who settled the baron's war, bolstered Henry's regime, and introduced Odoricus, whom he knew through Roman contacts in the 1250s. Westminster was “politically” papal in a way that rendered Canterbury irrelevant. Nor is there a shred of evidence that Odoricus or his team were available in the period 1230–1250 simply to re-lay an English pavement; yet at page 340 we find the confident assertion that the Cosmati mosaicists “arrived” in Canterbury. But why not English or French mosaicists? Once again, the author's unwillingness to consider the breadth of evidence for Opus Alexandrinum throughout the Latin and Greek worlds has placed them in an unsustainable position, uncorroborated by consideration of the history of the Cosmati companies (P. C. Claussen's excellent Magistri Doctissimi Romani [Stuttgart, 1987] is omitted). In effect the Cosmati are demoted. Yet the authors at no point raise the possibility, implicit in their own logic, that the Canterbury pavement actually post-dates the Cosmati works and was a response to them, rather than a preparation for them.

Given the high documentation and production values of this study, it is a pity to highlight such difficulties. But if close archaeological reading is to be durable it must surely be accompanied by a fully international horizon of understanding and a more sceptical tone than is apparent here. The present reviewer may certainly refer to this book in the future, but with caution.