Haunted by a boneless beast

To turn an octopus into a penis; to elongate it to new extremities; to make a wandering ghost where once there was a heroine: these seem like the acts of a Dali disciple or a drug-user in an aquarium. Yet the dictionary maker, that humdrum linguistic drudge, is allowed his moments of magic too.

As all readers of ancient Greek know well (most of us too well, for we consult it more than we would like), our standard dictionary in English is that of H. G. Liddell and Robert Scott, 19th-century Oxford classicists of whom the former is better known as the father of Lewis Carroll's Alice. Their work has been many times revised since 1843 (the current edition is the ninth and includes an extensive supplement), but it is still unsatisfactory in the eyes of its critics. Prominent among them is John Chadwick, whose contribution to the study of Greek through the decipherment of its earliest Linear B texts was a prime cause of current revisions to Liddell and Scott. What he now urges is a fundamental overhaul, taking into account modern information technology, the real requirement of readers (efficiency = usefulness V weight is his formula) and a strong dose of the imagination and reason which he feels that his predecessors have lacked.

This book is a set of sketches for that new dictionary which, because of the demands of time and money for its completion, the 77-year-old Chadwick suspects that he will never see. Lexicographica Graeca is a spirited and spikey work whose tone strongly belies the conservatism of its title. In selecting some 50 Greek words whose dictionary definitions are misleading in different ways, he hopes to signal how future reformers might follow in his path.

Chadwick's quick-changing octopus appears in a notoriously awkward passage from the poet Hesiod describing a winter day in the Boeotian countryside. This, he says, is the time when a mysterious boneless beast "gnaws his foot in his fireless house and wretched home". Generations of scholars have seen this creature as an octopus and duly seen their views reflected in Liddell and Scott.

But there are problems. Much of the archaic vocabulary here is obscure. What sort of poet would call an undersea lair a fireless house? The Greek phraseology also requires the emphasis to be on the beast eating its own foot in winter - as though on summer days it were its habit, by contrast, to nibble the toes of a friend. Other explanations have, therefore, appeared from time to time - that it was a cuttlefish or even a dinnerless sheepdog. In 1986 an Italian scholar even suggested that it was the male sexual organ which, according to Hesiod in a different passage, is more active in winter than in summer: instead of a sea-creature eating its foot, Hesiod would have meant a penis stretching to its extremity.

This is a classic problem to catch the scholastic mind. Rival critics cite modern marine biologists to show that octopuses do sometimes eat their own feet, anthropologists to show that primitive peoples prefer indirect language for animals, linguists to place Hesiod's line in the history of bucolic humour. Sex-in-the-cold or marine surrealism? The translator can make a choice. What, however, is the lexicographer to do? Should he accept the new, retain the hold or hedge his bets? A dictionary is supposed to be definitive, not a debating forum. Dictionary making is the scholarship of hard choices.

In Chadwick's view, Liddell and Scott and their followers have too often taken the easiest ways out. He is a harsh critic but he makes his points with brio. Many of the problems arise because of the shortage of texts for the earliest Greek period. According to the much-cherished "historical principle" of lexicography, the earliest use of a word is likely to reveal its essential meaning. Chadwick argues that Liddell and Scott will give a central meaning for a word based on wholly inadequate early evidence when the later usages show that a word had an original different sense.

The discovery of archaic words has become common in the 20th century. An issue close to Chadwick's heart (and close to the wallet of anyone who has to buy the rapidly expanding lexicon and its supplements) is the clogging of the work with newly discovered words of interest only to "a few epigraphists, papyrologists and linguists". Rather than applying tough standards for entry, the Greek lexicographers have held the door open to all-comers, creating a work which fails his efficiency = usefulness V weight test. This formula implies that an efficient dictionary weighing two kilograms ought to be twice as useful as one weighing one kilogram, "a condition rarely met by dictionaries which go on increasing in size, like Liddell and Scott".

Some of the extra verbiage, he argues, should be confined to electronic dictionaries in which weight is not an issue. But some words, he suggests, should not be in the dictionary at all. The desire to be inclusive leads to lazy thinking and the appearance of so-called "ghost words" which inhabit a lexicographical never-never land and would have been wholly unrecognisable to real ancient Greek speakers of any period.

He cites a word for "heroine" which crept into Liddell and Scott from a funerary inscription found in Sicily in 19 but which fails most other tests for being a proper part of the language. He suspects that a fundamental review would uncover many more ghosts sheltering among its densely paragraphed grey pages.

Where will these strictures lead? The history of dictionary making is spattered with new projects that never developed beyond their ABC. Once upon a time great individuals, Johnson, Littre, Grimm, Liddell and others, could set their own purposes and fulfil them: now huge teams and budgets are needed, with consequent vagaries of purpose that may often make matters worse.

Liddell and Scott's creation, for all its faults, is an extraordinary monument to British scholarship in the Victorian age. Any successor would inevitably be a matter for international cooperation-with all the attendant problems familiar from other academic fields. The discovery of lexicographical errors is certainly not a new game: whenever they were pointed out to Liddell in his lifetime he used to say, "Scott wrote that part".

This book is, of course, one more opportunity for a scholar to say, "if only I had written that part". But it is also more than that. Much of the pleasure in reading Lexicographica Graeca comes from the peculiar passion that dictionaries evoke in the scholastic mind. Open it at almost every point, and, however rusty one's command of Greek, the character of the impatient lexicographer leaps off the page, the man who believes that with just a little more logic and magic, hard thinking and literary imagination, the bones of a great language (as well as its boneless creatures) can be finally revealed in all their glory.

Peter Stothard is editor, The Times.