This volume is much more than a catalogue of the Uley curse tablets. T., the maestro of reading cursive script, has worked for many years on the assemblage of the Uley curses to produce this definitive study and discussion. It will undoubtedly be the go-to book for anyone wishing to understand cursing, literacy or popular religion in Roman Britain.

The first 80 pages contain an overall discussion of the curses, beginning with a survey of those found in Britain as a whole, along with a comprehensive bibliography in an appendix. These comparanda from Bath and elsewhere are drawn upon throughout the volume. The focus then narrows to Uley, starting with the nature of the tablets at the site, including those in lead but without any writing. These are assumed to be defixiones, as there are nail holes, presumably to attach them to some part of the temple structure. Those with writing are much more informative, and the discussion continues to formulas and the content of the texts. T. identifies a formula unique to curses from Britain, where a gift is made to the gods, for example half of a pair of pewter plates that had been stolen. If returned, the god would receive this gift as a votive offering. As such, they are not ‘pure’ curses in T.'s view, and they are dedicated to gods such as Mercury or Neptune, rather than the infernal powers. Some of the curses ‘give’ the thief to the gods, i.e. put them under the power of the deity.

Who that deity was, as far as Uley is concerned, is a little ambiguous. The sculpture from the site and most of the curses are to Mercury, but there are also Mars, Silvanus, even Mars Mercurius, and a new local theonym, Aruerius or Arueriacus. Many of the curses use the double formula, Mercurius Aruerius, implying a Celtic epithet, perhaps derived from proto-Celtic *arweros (‘host’ or ‘army’). This would link neatly with the miniature spears and other military material at the temple site, and perhaps also explain the ambiguity between Mars and Mercury on some of the curses.

Amongst the common words in the curses are templum and fanum. The former is more common than the latter, which gives us an insight into what the worshippers called the shrine building at Uley. Archaeologists today would classify it as a Romano-Celtic temple, and the word fanum is often used by French scholars to contrast this type of provincial religious building from the more well known, but in fact less ubiquitous classical temple style. What the Uley curses tell us is that both templum and fanum were applied to Romano-Celtic shrines.

The onomastics of the personal names also yield a mixture of ‘Celtic’ and ‘Roman’, with the proviso noted by T. that Celtic parents tended to name children using Celtic, and Roman parents likewise named children in Roman style, with little admixing. Given the late Roman date of the curses, this tradition in Celtic naming is noteworthy for its persistence long after the conquest.

Internal dating for the curses derives from the type of script in use. Old Roman Cursive and New Roman Cursive are both present, sometimes together on the same tablet. T. suggests that the poorly understood transitional period between the two scripts, 275–325 ce, is in question here, as only one curse is wholly in New Roman Cursive. A minority of the curses are in capitals or using Greek letters.

On to authorship, the topic that brings the discussion section to a close. As T. says (p. 74), ‘To find so many handwritten documents in the countryside, far from any urban or military site, unless they were the work of a few professional scribes, must mean that literacy was more widespread than the distribution of “monumental” stone inscriptions has suggested’. To be fair, as T. points out on page 308, the large Roman villa at Woodchester was only three miles away, which probably had a population the size of a smallish town, and the major urban centres of Gloucester and Cirencester were both less than twelve miles distant. That is a day's walk for pilgrims, who may have made special journeys to Uley for cursing. But were they literate? Some of the curses look like fair copies of an earlier draft, and some show ‘proof-corrections’ and crossings-out. Conversely, some are fluent and fill the space on the tablet to good effect, suggesting scribes who were used to the format and the unusual lead writing surface. But evidence of many curses by the same hand is not forthcoming, and each seems distinctive. Was it a ritual requirement that the curser, or their literate representative, wrote out the curse themselves, even if they were copying a formula written in draft? If so, the implications for the extent and depth of literacy in Roman Britain are significant. T. concludes his consideration of this important matter by referring back to the spelling variations in an earlier section of the discussion (p. 79), ‘The reason for this variation in spelling is surely that their authors did not read the word: they heard it spoken, and spoken at Uley. They must have asked someone what to say when they wrote to Mercury. Who better to advise them than the priest of Mercury? … for the practical purpose of discussing texts, this chapter may end as it began, by treating the petitioner himself as the author of his tablet’.

The bulk of the volume (pp. 84–361) consists of the catalogue entries, each with images, drawings, transcriptions, translations and commentary. There are 87 tablets from Uley, together with an additional seven from the nearby and similar assemblage of Tarlton. Some of the curses are now well known by being previously published in Britannia and elsewhere, along with their incorporation into many textbooks and exam questions. However, the full conspectus and associated apparatus is most welcome for its detail, richness and quality. This is a magnificent achievement by T., and it is surely a milestone in our understanding of society and religion in Roman Britain. The volume will repay many visits and, as a flavour, here is the translation of number 78 (p. 307):

As an immediate reaction to this breathless, astonishing document, a curser in the rural Roman Cotswolds has 100,000 denarii, apparently capable of being spared to dedicate to the deity. An eye-opening insight into late Roman Britain, indeed, as is the whole of the volume.