Computer generates Beirut's noble past

When this earthquake of might suddenly happened, the sea drew off and was poured back and restrained from the city of Berytus and from the rest of the cities of the seacoast of Phoenicia, and stood in a heap in its midst by the decree of God as far as a distance of two miles . . . (then) there came the great earthquake which overthrew the buildings of the cities, especially in Berytus. It brought down and crushed those who had escaped from the sea . . . the sea was lifted up over their heads behind them, and the earthquake struck the city.

Chronicles of Pseudo-Dionysius

JEAN-PIERRE Asmar can call up on his computer the ghosts of the earthquake that destroyed Roman Beirut. The screen slowly fills with a map of the streets and buildings destroyed in the country's 20th-century 15-year civil war, a more shameful and more bloody tragedy than the catastrophe that overwhelmed Beirut in 551 BC.

But when Asmar presses a new key, a map in red and green emerges beneath the modern-day city. A medieval wall and the site of two 'lost' Crusader castles creep in across rue Weygand and the modern-day port. And to the east of the castles, a red rectangle appears, beneath which, Asmar suspects, lie dozens of Roman columns swept down in the 6th century BC.

It is a curious sensation to watch the computer generate the past. It is as if the machine is digging beneath the civil war ruins of Beirut to the Ottoman, then the Roman then the Persian cities.

Asmar sits beneath stone arches in the basement of Beirut's restored Serail - headquarters for the Lebanese and foreign archaeologists working on four excavation sites in the devastated city centre - and admits, even after working for weeks on the maps, that the images continue to fascinate him.

On computer, some early streets align with present-day boulevards, others cut through the ruins of offices and shops. Frequently the medieval and modern streets are slightly out of 'true', suggesting that the early map, printed without modern technology, was inaccurate.

There was never any doubt of the treasures that must lie beneath the rubble of Beirut. Berytus, as the Romans called their great port city on the Phoenician coast, was a pearl of the Roman Empire, a powerful commercial centre and, later, the site of perhaps the most famous law school of antiquity.

Already Professor Leila Badr of the American University of Beirut has found a cluster of 5th-century BC votive figurines in an excavation site behind the wreckage of the shell- smashed Rivoli cinema, while other archaeologists have discovered a layer of sand over Roman pavements - quite probably the sand that was dumped on the city when the sea returned to kill its inhabitants in 551 BC. But what makes the Lebanese excavations unprecedented is Asmar's computer.

The computer image of Beirut's medieval walls was traced from a British map of 1841, brought to Lebanon by John Schofield, a Museum of London archaeologist temporarily funded by Unesco as a consultant to the Lebanese, who are excavating their capital.

Lebanon has been regularly invaded by foreign powers, but for the existence of this map, unearthed in the Public Record Office at Kew, it can thank the Royal Navy. It was the British who drove the Egyptians out of Beirut in September 1841 - they bombarded the city from warships - and it was a British naval map-maker called Davey who sketched the contours of the city at a time when it still boasted most of its medieval walls and two Crusader castles.

'We've never before done anything on this scale with a computer,' Schofield says. 'We are putting 160 hectares on to the computer - that's the most ambitious plan I've been involved in during my 20-year career. No one's ever had to develop an entire city before, so this is an extraordinary opportunity for archaeologists to advise where buildings might accommodate Roman or earlier remains and where the foundations of new buildings might be laid without injuring the archaeology beneath.

'We can plot sites from the 1841 map, then underlay the map of the modern city. Sometimes we have to 'stretch' the lines of old roads to make them fit - it's where Davey might have got his map wrong - and it's like . . . like mental plasticine,' he says with triumph.

A similar but much smaller scheme was undertaken in London when a computer was used to map the Roman amphitheatre in Guildhall Yard. But Schofield and his colleagues - as well as Lebanese, there are Syrians, French, Australians, Austrians and Italians working on the archaeological project - are trying to trace the Roman columns of Beirut with a piece of remarkable detective work.

The 1841 map shows the location of two Crusader keeps, one of them marked Bourj al-Masalla (Masalla's Tower) and the other identified as Bourj al-Mina (Harbour Tower). No Roman remains are obvious. But Schofield and his colleagues have noticed an intriguing detail in a sketch of Beirut by the British artist W H Bartlett in 1838, just three years before Davey drew his map. The picture shows the two castles - the nearer one linked by a bridge like the extant Crusader keep at Sidon - and, to the far right, a waterfront that appears to have been constructed with horizontal pillars. Schofield suspects they are fallen and re- used Roman columns.

From the Bartlett print, it is evident that the waterfront lay just to the east of the Bourj al-Mina. And by tracing this same harbour site on the 1841 map, Asmar is certain it lies beside the present-day port road, beneath a building so pulverised by shells it is to be demolished.

Computer technology has therefore pinpointed where archaeologists can dig for Roman columns and artifacts that could later adorn the archaeological park, part of the plan for Beirut's new city centre.

Asmar and Schofield hope soon to underlay a Roman map on to their computer. This map was drawn by the French archaeologist Jean Lauffray at the end of the Second World War, after a team of Frenchmen had sketched their own map of the Beirut of antiquity. At the end of the First World War, Lebanon had fallen under the French League of Nations mandate, and France proceeded to build a series of boulevards in the centre of Beirut; but as it did so, French archaeologists partially excavated the Roman city beneath, discovering the site of the Roman forum next to what was to become the Lebanese parliament; a few re-erected but bullet-scarred Roman columns still stand on a roundabout today.

The French map is partly conjecture; in her fine book on Beirut's archaeology, Beirut through the Ages, Nina Jidejian of the Beirut Museum points out that what the French thought was a Roman theatre turned out to be a Roman bath. But Schofield feels it could be incorporated into the computer maps. Lauffray's 1944-45 map could be grafted on to the 1841 map, then on to the present roads, giving the archaeologists further directions on where to dig for the city's treasures.

Until last year, I would regularly drive behind the Beirut port road, over what was clearly a Roman column buried in the mud surface of the highway. A few weeks ago, a group of Shia Muslim refugees living next to the road covered over the column with earth to ease the passage of drivers.

In only the past two months, the Lebanese have uncovered three delicate, late Roman, fish-scale-patterned mosaic floors in Martyrs' Square, as well as a Roman pavement behind the gutted Maronite Cathedral. A layer of sand - presumably from the tidal wave - was found beneath Ottoman buildings. Sometimes, of course, the romance of archaeology wears a little thin. A long-disused Ottoman drainage system discovered by archaeologists still smelled of sewage. But Leila Badr's diggers have unearthed a clay face mask, delicate votive figures of the 5th century BC, a beautiful model of a dog and a fractured but graceful Hellenistic pot.

The dead of antiquity were perhaps sending Beirut's civil war survivors a message when Badr's team found the pot. On the handle, a Greek inscription read: Peace and Happiness.

(Photographs and map omitted)