Old Greek at bargain prices

If you look up krokotos in Liddell and Scott, you will find an entry "saffron-coloured robe, worn by gay women". This is not as intriguing or contemporary as it sounds, being in fact an indication of period charm and a reminder that the lexicon is a Victorian edifice in need of constant maintenance (the word "gay" was deleted in the 1968 supplement). Another compelling reason for revision is the emergence in the 19th and 20th centuries of an immense amount of new data from papyri and inscriptions. The most extensive revision of the lexicon took 15 years, being completed in 1940 in the form of the ninth edition, commonly referred to as LSJ (Liddell-Scott-Jones). A supplement appeared in 1968. Now we have a revised supplement, twice the size of the original, representing the fruits of 13 years' work monitored by a committee of the British Academy, issued either bound with LSJ or as a separate volume.

One cannot properly review a lexicon without having had the use of it for several years. No reviewer notices or bothers to comment when a lexicon gets it right, being on the lookout for things that ought to be changed. On the other hand, the reviewer's task is made more difficult in this instance in that it is impossible to assess the level of performance in relation to aspiration. Anyone who has been involved in academic life in recent years will be heartily sick of expressions like "aims and objectives" and "mission statements", but with the revised supplement something of that nature would have been of help. There is no editor's preface. Instead we get five anonymous pages outlining the history of the project. In it we learn of a new policy regarding items written in the Cypriot syllabary and are offered an explanation of the methods of citing the newly introduced Mycenean entries (a welcome innovation). We also receive a health warning about the quality of etymological entries in LSJ. More alarmingly, we are also warned that many entries are in need of total revision. There is, however, no proclamation of overall policy or methodology. Exactly how much of LSJ has been corrected? If, as in fact soon becomes clear, only a select number of entries has been rewritten, what is the basis of the selection? One stated criterion is length. Shorter rather than longer articles have been recast, but we are not told how many and why. How complete is the supplement with regard to words omitted from the lexicon and first supplement? We are told that the editorial associate was responsible for most of the epigraphical material and we are given the names of various scholars who assisted with papyrus texts and made other contributions, but not told exactly what any of these individuals did. Was there a systematic reading and slipping of texts or did the revisers rely solely on published work, word indices and material submitted to them or already collected?

I have tested 100 cases where I knew the lexicon and the first supplement were in error and that this had already been pointed out. I find that the supplement has taken account of 44. In electoral terms 44 right against 56 wrong is a landslide defeat for accuracy and completeness, but in terms of revision of a work of this scale it does not seem to me all that bad a result. I sincerely hope that it does not underrepresent the supplement's failings.

But some omissions and errors surprise and disappoint, since the relevant information is so easily accessible in print. I miss beinema (the action noun cognate with the obscene verb bino) which first becomes known in 1968, is mentioned in several articles as well as in the Bulletin Epigraphique and is present in the Index des mots to that work. Then there is a non-word siteutorios: "poultry-fattener". The supplement does not know of the standard edition of the work from which the word is cited, the Philogelos. There we learn that the best manuscript has the expected form, siteutarios. Louis Robert, writing without knowledge of Thierfelder's edition, restores it by emendation.

The quality of lexicography in Liddell and Scott (especially the ninth edition) has come under severe fire from John Chadwick. I have to report at least one poor example of the lexicographer's art in the revised supplement. The entry which introduces a new meaning, "bury" for the verb ballo, is marred by the addition of "illicitly". This is interpretation of the context, rather than definition and is shown to be wrong by one of the inscriptions attesting the meaning, which runs "let no other person be put into my tomb ... except me". What we have here is a characteristic example of the semantic convergence of verbs meaning "throw" and "put": ballo takes over one of the functions of the principal verb of placing in Greek, tithemi.

One devoutly hopes that the publication of this stopgap and (considering the magnitude of the task) under-resourced volume will not see the end of supplementation and correction of LSJ. The preface makes it clear that the computer has up to now played very little part in the revision, which these days seems absurd. Here is surely where the future of the lexicon lies. Materials accumulated for the project of revision should be made accessible to computer-literate scholars with a view to the continual supply of addenda and corrigenda.

Priced at Pounds 100, the lexicon bound with the revised supplement is an outstanding bargain, especially tempting for those who do not already possess the lexicon or for those like me whose copy is, through constant use, somewhat the worse for wear. It goes without saying that Liddell and Scott is still indispensable for all serious students of Greek language and literature.

David Bain is professor of Greek, University of Manchester.