As the world's oldest creative writing, Sumerian literature has a special place. Nor is it a small place. Somewhere between three hundred to four hundred compositions, have come down to us, typically varying in length from 50 to 800 lines. They are known to us from cuneiform texts recovered over the previous two centuries. The majority of manuscripts date to the Old Babylonian period (nineteenth/eighteenth centuries bc) but the compositions which they record were overwhelmingly composed earlier - direct evidence takes us back to the Ur III empire and the Early Dynastic period and there is no telling just how much older some of the compositions may truly be. Their origins are lost in hoary antiquity. But the textual documentation is vast. Most works are preserved in multiple manuscripts, in some cases the duplicates running into the hundreds, and the textual situation which these manuscripts present is highly complex. The corpus is characterised by a very large number of divergent writings, orthographic variants which demonstrate not simply errors on the part of the scribe - though these do, of course, feature - but rather attest to the fact that Sumerian literature, in the form in which we have it, allowed and indeed perhaps encouraged a degree of parallel traditions in the manifest text(s). Coming to grips with how to deal with this is at the core of the present volume. It aims to formulate the basic principals for the textual criticism of the Sumerian manuscript tradition. Delnero starts with a consideration of the approaches to textual criticism developed for the study of Greek, Latin and Hebrew texts, and rapidly confronts us with the fact that the principals used there are not wholly applicable to Sumerian literature. For contrary to the cases with the classical and biblical compositions, which are predicated on the reconstruction of a hypothetical single original text, with Sumerian literature it is assumed there was, generally, no such thing. This poses some problems for the works which rely on standard editions - for example dictionaries (notably the Pennsylvania Sumerian Dictionary, PSD) and electronic resources (notably the Electronic Text Corpus of Sumerian Literature, ETCSL): these projects work from composite texts created from the sources available but, almost inevitably, generate standardised editions which do not actually fully correspond to any tradition from antiquity. In other words, for most of these compositions final texts do not and never did exist. A useful model from more recent times is provided by certain Latin compositions from medieval France. It is one thing to correct a scribal error, quite another to compress legitimate parallel traditions into one single modern edition. At stake, then, is the distinguishing of intentional from unintentional variants and the fact that the ways of approaching these are different.

Delnero formulates a classification of the types of variants found: scribal errors, regional and temporal variants, orthographic variants and other idiosyncrasies to bring order to the whole question. To make the task manageable and achievable he concentrates on the ‘Decad’ (Tinney's term) of works comprising royal and divine hymns, mythological compositions and songs which appears to have been at the core of the Old Babylonian school system. A particular advantage of this approach is that it allows for a systematic recognition of the range of variants at large and, by statistical documentation, to determine which are more or less common and how they are to be understood.

Turning to formal classification, what Delnero calls ‘mechanical’, (ie scribal) errors fall into three principal categories: errors made when copying form another manuscript, errors made in making a copy from dictation and errors made in writing down a composition from memory. At this point it should be mentioned that the majority of Sumerian literary manuscripts - for some compositions indeed all the exemplars - are school texts written by trainee scribes in the course of their education. These scribal exercises are immediately recognisable by their format (Delnero follows the typology proposed by Civil) and the recognition of this fact furnishes some of the reason for the multiplicity of variant writings. Scribal errors do indeed form a significant proportion and all the basic types - omissions, haplography, dittography, phonological errors, visual errors - are represented.

Local and regional variants relate to alternative writings which diverge from the ‘standard’ and which are found two or more times in manuscripts from sites other than Nippur, that city being taken as the standard as that is where the greatest number of Sumerian literary manuscripts hail from. The Enlil temple may indeed have served as the de facto national repository for Sumerian literature, in effect the Académie Sumerienne of its time. The other cities which have yielded sufficient manuscripts to offer the possibility of mapping geographical variants are Ur, Kish, Isin, Sippar, Babylon, Susa, Uruk, Larsa and Me-Turan (Tell Haddad). Nevertheless, the clear identification of authentic variants from these places is hampered by the facts that in many cases the exact provenance of the tablets was not recorded, that they may in consequence differ in date, that the number of texts is too small, that many are too fragmentary, and that there are too few instances where sites have yielded multiple duplicates of the same texts. In practice the site where an elucidation of place-specific variants can be most convincingly demonstrated is Ur.

With diachronic variation there is a different set of considerations. The ancient recording of Sumerian literature spans over two millennia. Comparing manuscripts and versions from across this time period throws up issues of the oral origins and probable continuing oral traditions of the repertoire; divergences due to developments in grammar, orthography and pronunciation; substitutions resulting from semantic differences in the thesauri of different time periods; modernisations carried out by scribes in order to bring texts up to date and compliant with contemporary practice; and the existence of different recensional traditions. This said, diachronic variation can be studied, particularly with regard to the manuscripts excavated at Nippur where, apart from the main body of Old Babylonian copies, there are also texts from the Early Dynastic, Ur III and Middle Babylonian periods.

The next classes of variants chiefly relate to the work of trainee scribes - variants within sources compiled by the same scribe (or group of scribes), idiosyncratic variants (erroneous omissions, phonetic spellings and grammatical errors committed by individual scribes) and interpretative variants (resulting from a scribe who has misinterpreted a passage and introduced an incorrect substitution).

If the above is somewhat bewildering, it correctly conveys the fact that carrying out textual criticism on Sumerian literary manuscripts is no straight-forward task. Turning to the issue of how to formulate a procedure for evaluating textual variation, Delnero points out that the first step must be to try and classify the nature of each variation, and proposes a methodology for doing so involving the characteristic statistical signature of each type of variation in terms of the internal and external patterns of occurrences.

The approach and methodology developed by Delnero represent a giant leap forward. His command of the material and the issues involved is genuinely impressive. There can be no doubt that this is a work of outstanding importance. As a striking parallel, it is interesting to note that while Delnero was working on his approach to the textual criticism of Sumerian literature, Martin Worthington was formulating analogous approaches to literature in Akkadian. Worthington's Principles of Akkadian Textual Criticism, published in 2012, is an equally ground-breaking work which will transform the field. Together these works set new standards for the textual criticism of Mesopotamian literature. To return to the work in hand, Delnero deserves the greatest praise for an exceptional work which is a major contribution and which will clearly guide future phases of Sumerian research for decades to come.