This book really does fill a long-felt need. It is a comprehensive account of a key artefact type of Migration-period Europe, the cruciform brooch. Cruciform brooches developed in the early fifth century ad in north Germany as one of many variants on the bow brooch of Iron Age and Roman Europe, a small copper-alloy brooch characterised by a head plate with three knobs (hence ‘cruciform’), a bow and a narrow foot ending in a zoomorphic terminal usually described as a horse’s head. Later versions, found especially in Norway and England, are much larger and have elaborate decoration.

Despite attracting attention from every scholar who has worked in this period there has not been a recent publication focused on the English cruciform brooch, so this work is very welcome. The first part of the book presents a typology and chronology of cruciform brooches, with supporting catalogue and illustrations, while the second part analyses and discusses their significance. Detailed artefact-focused research is combined with a theoretically informed analysis of the role played by cruciform brooches in early Anglo-Saxon society. It is an excellent example of how valuable this kind of research remains. The book is attractively produced and well-illustrated.

Martin identifies four main types of cruciform brooch (with many sub-types), which in turn are assigned to three chronological phases. In outline the small simple examples, type 1, are early, the large elaborate ones, type 4, are late, sandwiching the large number of less easily sequenced types 2 and 3 in the middle. This confirms, modifies and greatly expands previous schemes. Maps of the three phases (figs 49–51) show a dense concentration in East Anglia and Lincolnshire, especially in phases A and B, with a more dispersed distribution for the latest, phase C types. Few cruciform brooches have been found south of the Thames, apart from a small distinctive group in Kent in phases A and B. Absolute dates run from the early fifth to the mid-sixth centuries ad (table 12, p 128). The sequence of types is explained convincingly in terms of development from a locally varied dress fastener to a large and elaborate, but more standardised, brooch interpreted as symbolic of elite identity. The concentrated eastern distribution of the earlier brooches supports other research in suggesting that this region saw its most intensive settlement during the fifth century ad. The traditional identification of this region and the artefacts as ‘Anglian’ does reflect an identity created, however, within England, thus reflecting ideas about ancestry rather than being an actual package of ‘Anglian’ material culture imported from Angeln.

In later chapters Martin discusses the manufacture, use and repair of the brooches; their association with selected older women, and above all, their significance as ‘bearers of tradition’. In another paper Martin pointed out the oddness of ornament on female adornments being of predatory beasts or male humans (Martin Reference Martin2013) and in this book he associates the brooches, and their intended meanings, with ‘the movers and shakers of the Anglo-Saxon world’ (p 189). Does this mean the women played a direct role – or were they simply wearing their husband’s wealth and political authority?

Even in a scholarly work of this standard corners have occasionally been cut. The database does not include all the very large number of cruciform brooches recovered in Norfolk by metal detectorists. Martin’s sample includes 2,075 brooches from the whole of England (p 9), while from Norfolk alone more than 1,000 had been recorded by 2009 (Chester-Kadwell, pers comm, post-2009). The reason for this discrepancy is that the Norfolk finds are not all recorded online – they are in an archive that researchers need to visit. It is ironic that the county which pioneered the recording of metal-detected finds, and which has an unrivalled record of them, should be under-represented in research like this. Dependence on sources available online is likely to increase, and with it the possibility of distortion. Martin acknowledges the lack of Norfolk finds, but only in a footnote (p 9). The unwary reader may take his numbers as definitive. Martin’s maps do in fact show a dense distribution in Norfolk, and are not very different from Chester-Kadwell’s, but clearly an even darker density shading would have been appropriate. He seems not to have used the most recent publication of Spong Hill, so there are discrepancies in numbers and percentages between his numbers and ours (for example, Hills and Lucy Reference Hills and Lucy2013, 32–3, table 2.1, vs Martin 2015, 224, table 17).

These criticisms do not affect the conclusions presented in the book, nor do they detract from admiration for its achievement.