A dusty loft in Oxford's strangest college, All Souls, is where the whacky but worthwhile Lexicon of Greek Personal Names was born more than 20 years ago. Things have moved on and upwards since then.

Today, the project to collect and publish all known Greek personal names from the eighth century BC to the mid-seventh century AD is housed in the rather grand Clarendon Building, next door to the Bodleian Library and the Sheldonian Theatre. But it continues to enjoy a privileged status as one of the British Academy's most prestigious projects.

There are around 30 long-term research enterprises which have been accorded the hallowed "Academy Research Project" status and which attract almost automatic annual funding.

This year, the lexicon project has received Pounds 57,515 precisely. Next year, it will get Pounds 65,500 - less than the medieval Latin dictionary project (Pounds 81,540), more than the prosopography of the Byzantine Empire project (Pounds 64,285), and substantially more than the projects labelled "Oxyrhynchus papyri" (Pounds 36,000) and "Corpus Signorum Imperii Romani" (Pounds 4,250).

ÁñÁ«ÊÓƵ

This funding supports one of the editors, Elaine Matthews, and a couple of assistants. The project's founding editor, Peter Marshall Fraser, a former warden of All Souls, is unpaid. Now 77, Mr Fraser, a dashing silver-haired don who helped the Greek resistance in the second world war and won the Military Cross, is one of the country's leading classical epigraphical scholars. In other words, find an inscription, a citation, a motto in any of the three Greek dialects or umpteen regional alphabets, and he will decipher it and all but tell you what the writer had for breakfast.

Down the years, many people have found Greek inscriptions, especially those venturing on the traditional grand tour in the 18th and 19th centuries, and by the turn of this century most of the thousands of Greek inscriptions that were ever going to be found had been found and had been published. These collections, Inscriptiones Graecorum, known as IG for short, form the primary resource of the lexicon - or rather the Greek onomastikon, to use the jargon.

ÁñÁ«ÊÓƵ

An earlier onomastikon had been prepared in the 19th century by two German scholars, Pape and Benseler. Based on literary sources, it was published prior to the explosion of new collections of inscriptions in the late 1800s, and it quickly became dated. As Mrs Matthews puts it: "It was like The Forsyte Saga coming out in black and white, just before the arrival of colour television."

The inscriptions appeared in many different places: buildings, coins, vases, tombstones. The lexicon project has deciphered around 500,000 personal names - in effect, people - and beyond this difficult enough task, it has endeavoured to discover when and where they lived. This has been achieved by interpreting such evidence as the shape of the letters, the artistic style of the tombstones, and even the formula, used to construct the inscription. As Mr Fraser explains: "If it's a Christian formulae, you immediately know where you are."

Generally speaking, ancient Greeks possessed just one name - usually the name of a god, but sometimes a river or a mountain, and sometimes a character in a ÁñÁ«ÊÓƵric epic. In Athens, the most popular name was Dionysius, after the god of wine. According to the computer database, 1,103 people of the 62,360 people who have been identified over a 1,400-year period shared this name. Other popular names were Apollonius and Aphrodisius. The cynic, mindful of the Pounds 60,000-plus annual price tag, might say, so what? The answer lies in the importance of personal names as a key to understanding Greek society, something first appreciated by the French scholar J. A. Letronne in 1845.

According to Mrs Matthews: "He appreciated that personal names reflect the language, the landscape, the cults and the institutions of past societies - especially since, by then, literature had been soaked dry of information."

ÁñÁ«ÊÓƵ

Mr Fraser says the lexicon will prove useful to scholars because, among other things, it will document the spread of religion as attested by the adoption of theophanic names, the existence of social class as determined by different names and the demographical distribution of the Greek world.

Already two volumes have been published, and there are plans for four more, to be completed by 2005. But these would only cover what is being described as "the first series". There are tentative plans for "a second series", to cover the ancient Greeks who lived in Egypt, Syria, Cappadocia and beyond. It could take a lifetime.