To those who have been following Barrett's meticulous research on the origins of printing in China, which has been laid out in a series of detailed articles over the last decade, the title of this book will come as no surprise, startling though it may be to other readers. Unlike the articles, which are mostly addressed to an audience of specialists, The Woman who Discovered Printing presents a more accessible synthesis of his work to date as well as incorporating the results of recent research. Barrett makes it disarmingly clear however, that this book is definitely not the last word on the subject. Further dramatic archaeological discoveries and/or the identification of documentary evidence that has hitherto been overlooked may well enable Barrett in a few years’ time to fill in some of the gaps. For the present, however, all that can be done is to assemble the available evidence, some of it discovered by Barrett himself, and to speculate on possible scenarios and explanations, and nobody is better equipped than Barrett, in terms of knowledge of the times and the religious context, to do that. He is careful enough, however, to alert us to the speculative elements that underlie all historical writing. Perhaps that is why he begins with an avowedly imaginary glimpse of the Venerable Bede gazing at contemporary events in China. But this conceit also serves to remind the reader of the comparative enormity of the scale of manuscript production in Tang China, to say nothing of the discovery of printing.

In a nutshell, Barrett's tentative argument is that printing was first undertaken to produce large numbers of copies of a text in China, some time around the year 700, by the woman emperor Wu in conjunction with Fazang, a Buddhist monk of Sogdian descent. The circumstantial evidence is compelling, but it would not yet stand up in court. One of the many frustrating problems Barrett discusses in this scrupulously honest book is the lack of straightforward contemporary references to printing and the lack of any sure evidence of printing texts until a century or so later. This might seem fatal to Barrett's conclusions, but as it happens it is not. The reason for this is that an example of printing was found in 1966 in Korea that appears to have been produced before 751 and many hundreds of examples produced in Japan in the 760s survive to this day. The texts printed in both Korea and Japan are all from the same sutra of magical spells which was translated into Chinese at the end of the seventh century, and, although some diehard nationalists would object, it is surely absurd to suppose that the printing of texts was developed independently in either Korea or Japan. For this very reason, indeed, it has been claimed that the example from Korea was actually printed in China and merely exported to Korea. Be that as it may, the obvious source, not only of the Chinese text itself but also of the means used to reproduce it, was China, with which both Korea and Japan had close connections at the time, and few Korean or Japanese scholars dissent from this view. Therefore, irrespective of the lack of evidence from China, it must surely be supposed that printing had been practised in China by the early eighth century at the latest.

There is also another sense in which the lack of evidence is not fatal, and that relates to the evidence for printing in Korea and Japan. Nothing had been known of printing in eighth-century Korea until the discovery made in 1966 and the account in the Japanese official history refers to the mass production of texts but does not trouble to mention that they had been printed. Furthermore, in both cases, as in China, there is little sign if any that a major technological advance, as it strikes us today, was put to any use at all for decades if not centuries. Barrett advances various possible explanations for this, including the ability of manuscript production to meet the demands of the day and aristocratic disdain for publication. He also draws attention to the wholesale manipulation of the historical record after Wu's death, and this is surely what in the Japanese case explains the silence in the historical record on the costly and technologically demanding undertaking to produce a million printed spells, and a million miniature pagodas in which to contain them, in the years 764–770. In the Japanese case, too, the problem was a woman sovereign, Shōtoku, whose reign had aroused great hostility, just as had the reign of Wu. Is it farfetched to suppose, in the aftermath of these two controversial reigns, not only that the historical record was altered but also that the technology they had championed was too dangerous to play with? At any event, it was indeed centuries before printing was undertaken again in Japan and many decades in China, so far as we know from the historical record and surviving fragments of printed text.

Barrett's account ranges widely, from paper production at one end of the process to silent reading at the other, but what he sees as the context for the ‘discovery’ of printing in China is religion. He is therefore more concerned to explicate the desire to reproduce texts than to elucidate the history of related technologies such as pattern-printing on textiles, though he does not ignore these either. What ‘religion’ means here is the shifting climate of popular and elite Buddhist and Taoist practices, for motives for the replication of texts can be found in both traditions. There is no single ‘explanation’ to be found here, though: we need to keep in mind, Barrett argues, such diverse factors as the replication of texts as a ritual act, the compelling need for accuracy in reproduced texts and above all the function of words and texts as a relic in Buddhism. But if the examples found in Korea and Japan and the evidence Barrett has found for China are any guide, it seems incontrovertible that the earliest printed texts in China were undertaken as a ritual act and were not produced for anybody to read: mere reading was easily satisfied by manuscript production. In the case of Japan, we know this for a fact, since the printed spells were placed inside miniature pagodas and then forgotten. The wellsprings of printing in East Asia were thus “spectacularly unlike the spread of printing in Europe” (p. 130), so to understand them requires ridding ourselves of ‘commonsense’ notions derived from European practices.

This is an unfinished story. There will be more to come in the next couple of years, from Barrett and from others, but for now this is an important, elegantly argued and very stimulating contribution to the debate on the origins of printing in China.