Tremendous advances have been made in the last three decades in the study of the language and history of Bactria, following the sensational discovery of close to two hundred documents from the 4^th to 8^th centuries ce. The corpus has been published by Nicholas Sims-Williams (Bactrian Documents from Northern Afghanistan, 3 vols., 2001–12), whilst François de Blois has made important contributions, in the light of these new documents, to the study of Bactrian chronology.

Although entitled ‘Studies’, this volume is not a mere collection of articles, but an attempt to carry out a comprehensive analysis of the dating and chronology of the Bactrian documents and inscriptions. It also has the added value of presenting a concise overview of the entire Bactrian corpus. Although Bactrian chronology (not to speak of its language and history) remains full of uncertainties, this volume demonstrates just how much can be inferred from a limited body of evidence on the basis of a rigorous philological methodology. The authors must be commended for their perseverance, precision, and accuracy, and for producing a relatively short volume that is densely written, rich in information, and very largely convincing.

The first part of the book is on the dated documents. There are chapters on dates and numerals, on the Bactrian calendar (which, it is argued, was closely related to other Iranian calendars and based on the same structure as the Zoroastrian calendar), and on Bactrian chronology, which the authors argue was based on an era of Ardashir, beginning in October 223ce. The Bactrian era has been highly contentious in recent scholarship; the authors give a fair voice to competing opinions, but make in conclusion a very convincing case. The second part of the book attempts, in great detail and very systematically, to date the undated documents on the basis of prosopographical, palaeographical, linguistic, and orthographic criteria. Allowance is made for geographical variation (complicated by the fact that nearly the entire corpus of documents is unprovenanced) and other possible variables. In all cases, the evidence is carefully assessed and clearly presented, enabling the reader to evaluate for him/herself the merits of each argument, and “to check its credibility in the light of new arguments or new material”.

The volume includes tables, maps, and comprehensive lists of the sources and their proposed dates. There are several appendices, notably on the Tochi valley inscriptions (first published in the 1960s; their meagre chronological elements provided an important starting point for the analysis of the more recently discovered documents), and on an attempted radiocarbon dating of a sample document. Additionally, there are short essays on the Achaemenid origins of the Zoroastrian calendar (in the light of an ancient Bactrian document published by Naveh and Shaked), and on calendar years and tax years in early Islamic Bactria.

A few small details:

pp. 21–23: here the authors discuss the dating formula of ‘days elapsed’, which is attested in Bactrian but also in Old Persian and other Iranian languages. They suggest, on the basis of Arabic usage, that this phrase is an archaism going back to a period when the date was given according to the number of elapsed nights in the month. This seems to me excessively hypothetical, and unnecessary. I do not consider it such a problem, linguistically, that the current day is included in the number of days of the month that have ‘elapsed’. This reasonably well attested usage, over many centuries and in a number of cognate languages, does not need to be explained as coming from something else.

p. 31: the tabular equivalent of 13 Jumada I, 243ah, according to the authors, is Tuesday 7 September 857ce; yet the given day is Friday. It is highly implausible that the month was delayed by three days because of weather conditions. New moon sighting is often delayed by one day, but rarely by two days, and nearly never by three. Moreover, poor visibility due to bad weather is unusual in August. To make things worse, according to the programme Kairos (which the authors have used), the new moon should have been visible one day earlier, hence 13 Jumada I would have fallen on Monday. The three- or four-day discrepancy must rather be due to an error in the date.

pp. 41–42: a Bactrian era from the beginning of Ardashir's reign is plausible, in my view, regardless of when Bactria was conquered by the Sasanians. The earliest attestation of this era is from year 35 (IKTB 10), thus according to the authors, 257/8ce, by which time Sasanian rule was long established in Bactria. We know other eras that were retrospectively established, for instance the regnal years of Seleucus I (and hence the Seleucid era) in late 4^th-century bce Babylonia, or more pertinently, the Hijri era.

pp. 109–112: de Blois’ argument regarding the origins of the Zoroastrian calendar is commendable, but its institution cannot be dated specifically to the reign of Xerxes, or to 481–479bce. As I have argued elsewhere (Calendars in Antiquity, 2012, pp. 177–178), such a precise dating depends on the assumption that the founders of this calendar were capable of calculating the vernal equinox to the same level of precision as modern astronomers; and moreover, that they waited a long time, possibly decades, for the equinox to coincide exactly with the first day of an Egyptian month (which only occurs during one four-year run in 120 years) before launching their new calendar. This is not how calendars are instituted. The Zoroastrian calendar was instituted whenever the decision to do so was taken – I would venture, not too long after Cambyses’ conquest of Egypt, but anyway at some point in the ensuing hundred years – taking as first month the Egyptian month that fell closest, at the time, to the vernal equinox.

I also consider it unlikely that this calendar was first instituted as a “purely religious” calendar, only to become “secularized” later on (as argued on p. 112). Zoroastrian priests had no reason to replace their traditional religious calendar (presumably, lunisolar) in favour of a new model from distant Egypt. The calendar now known as ‘Zoroastrian’ was most likely civil from the outset: it was appropriated by the Achaemenid kings as a trophy of their conquest of Egypt, and also, perhaps, as a stable, predictable calendar, well suited for administrative use across a very wide empire. This is not contradicted by the fact that its days were named after the Zoroastrian yazatas, and that the calendar was soon identified as Zoroastrian. Civil and religious functions are obviously not mutually exclusive (see Calendars in Antiquity, pp. 184–191).

All this takes us a long way, however, from Bactria of the 4^th-8^th centuries, on which this volume is a mine of information, and a most impressive feat.