Hesychius’ lexicon is undoubtedly the most important Greek dictionary to survive from antiquity. It provides the only evidence for the existence of a large number of rare words as well as crucial information on the meanings, dialects, and usage of others, not to mention numerous citations of lost classical literature. It is also very difficult to edit, being a huge work surviving only in a single, highly problematic manuscript. Kurt Latte began work on his edition as a young man in 1914 and left only two volumes when he died half a century later (Α-Δ published in 1953 and Ε-Ο published posthumously in 1966). Peter Hansen then spent a further 17 years on a third volume (Π-Σ, 2005), and the fourth was finally completed by Hansen and Ian Cunningham in 2009. A fifth volume consisting of indices had been promised, but after Hansen’s death Cunningham opted instead to revise Latte’s work, resulting in the work under review here: this is a revision by Cunningham of Latte’s 1953 volume.
Latte’s edition was pretty good in terms of the text it provided, the accuracy of its marginal attributions of particular glosses to different sources, and the usefulness of its detailed introduction, but it had a major drawback in the form of an inadequate apparatus criticus. Latte reported the manuscript’s readings only selectively, sometimes misrepresented them, and never reported their accents; one reason his readings were poor was that he relied on photographs rather than looking at the manuscript itself. Cunningham set out to address these problems and accordingly has produced an edition based on personal inspection of the manuscript (a complete re-collation), in which all its readings are reported in the apparatus with all their diacritics. Therefore nearly every entry in the apparatus is different from the corresponding entry in Latte’s apparatus, if only because the Greek words now have accents. In a sample of the apparatus for 200 entries (β 1-100, δ 1-100), 26 have changes to the manuscript readings going beyond accentuation (two changes to word division and 24 to the letters themselves)—and even where the readings are unchanged they are now more useful because one can have far more confidence that they are right. Of course, not having seen the manuscript myself, I cannot swear that Cunningham’s readings are always better than Latte’s, but given Cunningham’s track record I would be inclined to trust his readings.1
The apparatus has also been expanded and updated with other types of information, such as the sources of emendations. The apparatus of parallel passages has likewise been expanded and updated. They now offer more and better information, but unfortunately that information is often harder to understand than it used to be. For example, entries γ 19-22 are words beginning γαδ- where the γ represents a digamma, a fact first noted by the seventeenth-century Dutch scholar Daniel Heinsius; Latte’s apparatus here reads ‘ϝαδ-’, while Cunningham’s has ‘ϝ agn. Heins.’.
Changes to the text itself are not infrequent. In the same 200-entry sample, eight entries have changes to the actual text, i.e. the letters of the Greek. Five of these are clear improvements: two correct Latte’s typographical errors (β 90, δ 52), two print a corrupt manuscript reading (marking it as corrupt) where Latte emended into an implausible hapax legomenon (β 74, β 90), and one removes material not in the Hesychius manuscript, which Latte had added from another source (β 49). The other three changes to the text (β 12, β 27, β 87) are not necessarily improvements but do not make things any worse. Additionally, Cunningham divides two entries (β 87, δ 33) into multiple entries each, since originally separate notes had become mixed up with each other in transmission; these are major improvements. Then there are seven entries where Cunningham improves Latte’s punctuation, seven where he changes the marginal sigla indicating the source of an entry, many where he adds, removes, or updates references to literary works, and very many where he alters the in-text sigla—with the result that fully two-thirds of the entries in this sample have some sort of change to the text presented in Latte’s edition. With this level of alteration the volume is really a new edition rather than a revision; Cunningham would have been justified in removing Latte’s name from the title page.
The changes to the text and apparatuses are clearly a net improvement, but the same cannot be said for the changes to the extensive explanatory material also included in Latte’s first volume. Latte’s 45-page Prolegomena explaining Hesychius’ work, its complex history, and the difficulties of its transmission was the best introduction to Hesychius ever published 2 and had just been updated by Alpers (Hesychius vol. 3 pp. xv-xxiii); it is shocking to find it removed and replaced with a tiny and infinitely less helpful version.
The loss of Latte’s list of abbreviations is almost as serious—for the abbreviations themselves are still used freely in the edition. Entries from the Cyrillus lexicon are marked with an asterisk, and five different kinds of brackets are used in the text, but the reader is never told what these symbols mean. The margins are festooned with indications of the sources of individual glosses, most often ‘D’ (Diogenianus, a lexicon compiled in the second century AD and now lost except insofar as it was incorporated into Hesychius’ work in the fifth or sixth century) but also ‘K’ (Cyrillus, a late antique lexicon), ‘Ap.S.’ (Apollonius Sophista’s lexicon of Homeric words), ‘Att.’ (Atticist lexica), ‘On. sacr.’ (collections of scriptural names), ‘Orth.’ (Byzantine orthographical works), etc. Only ‘K’ appears in the list of sigla and only ‘Ap. S.’ in the bibliography; the Prolegomena has information that could lead to decoding several others, but it does not have any of these abbreviations as such, and it has no information at all on quite a few of them. Fortunately almost all the abbreviations can be found in volume 3 (pp. xxv-xxxiii) and the symbols in volume 4 (pp. xvi-xix), but how many readers of volume 1 of an edition will think of looking in volumes 3 and 4 for help in decoding the abbreviations and symbols? At very least a prominent cross-reference should have been provided.
This volume also greatly expands the sigla used within the text itself to indicate which parts of each Cyrillus-derived entry can be found in which Cyrillus manuscripts. This offers scholars more and better information, but at the cost of giving the text a more cluttered and intimidating appearance. For example, entry α 8697 now reads ‘*ἄφιξις· ἔφοδος vgαφο4 (AS19Br1189). παρουσία gαφο4(AS19)’. Scholars wanting to use this extra information are going to have to work hard—although v, g, A, S, and Br are in the list of sigla, the key to the superscripts can only be found in a footnote in volume 3 (p. xxv). (The superscript numbers are sequential numbers for entries beginning with α in each manuscript; superscript letters indicate what the first three letters of each entry are in that manuscript, if they are not the same as in Hesychius.) The non-specialist reader may well wonder whether this information is important enough to be worth cluttering up the text of Hesychius to this extent, and if it is so important, why the key to it has not been given in a more prominent location.
The other type of information in the text is references to the literary sources from which many entries seem to derive. These have been checked for accuracy and updated with references to new editions, which is very welcome—but it is a great pity that no attempt was made to address the general opacity of the abbreviation system used for these references. All the abbreviations used here are specifically excluded from the bibliography, and yet some are far from being familiar (e.g. ‘Dosiad. Ara 5’ = Dosiades, Βωμός line 5 (in Powell’s Collectanea Alexandria); ‘Ps. Sal. 16, 4’ = Septuagint, Psalmi Salomonis 16:4; ‘Cyr. hom. div. 14, PG 77, 1072’ = Cyril of Alexandria, Homilia diversa 14, in Migne’s Patrologia Graeca vol. 77 p. 1072). Some conventions will also be unfamiliar to most readers, such as putting two dots after a reference to indicate that other examples can be found in the same author. It does not help that the same abbreviation can have different meanings in different places, such as ‘Prov.’ when it occurs in the text referring to the Biblical book Proverbs but when it occurs in the margin referring to ancient paroemiographical collections, or ‘K’ in the text referring to Iliad book 10 but in the margin to the Cyrillus glossary.
On the positive side, Cunningham has made the reader’s life much easier by printing the numbers of all entries in full rather than continuing Latte’s confusing system of printing only the last two digits of each number—and his decision to retain Latte’s numeration avoids making the reader’s life harder. The book is meticulously proofread; I found only a very few typographical errors, none of them serious.
In short, this edition is an enormous improvement on Latte for scholars who have the prior knowledge needed to understand the extra information provided, as well as the motivation and interests to use that information. But there are not many such scholars, and for the average user the edition is at best a mixed blessing, as the benefits of a small number of improved readings in the text do not compensate for the loss of the Prolegomena and list of abbreviations, not to mention the increased difficulty of reading both text and apparatus. One hopes that libraries will keep Latte’s original volume along with this one, to ease readers’ difficulties, but that hope is unlikely to be realized; therefore this publication may in the long run damage the study of Hesychius by making the work more difficult to approach. Twenty-first-century scholarship ought to make important texts more accessible, not less: if we want future generations to be able to use the ancient texts we treasure and to appreciate our research on them, we need to break down barriers to understanding rather than building them up.
Cunningham is probably already working on a revision of Latte’s second volume, and I have every confidence that it will offer as many improvements to the text and apparatus as the current volume does. I hope it will also, before it is too late, make up for the deficiencies of this volume by providing a complete list of abbreviations and restoring Latte’s Prolegomena—ideally in an English translation incorporating Alpers’ changes from volume 3 and Cunningham’s own Prolegomena from volume 4, but failing that at least reprinting the original Latin, which is exceptionally clear and engaging.
1. Note especially his edition of the Συναγωγὴ λέξεων χρησίμων (Berlin: De Gruyter 2003).
2. Cf. Dickey, Ancient Greek Scholarship (New York: OUP 2007) p. 89, and note Alpers’ assessment (Hesychius vol. 3 p. xv), ‘Latte’s Prolegomena are of great importance not only for the Lexicon of Hesychius itself, but ... also for Greek lexicography in late antiquity and the Byzantine period’. Alpers’ corrigenda to the Prolegomena will of course lose much of their value once the work to which they refer is no longer available, as they do not make much sense on their own. And the general coherence of the whole Hesychius edition is impaired by having volume 3 begin with a set of corrigenda to something that is no longer there.