IELTS Academic Reading Practice Test 241

Petra lies in southern Jordan, in a valley among the sandstone hills that rise between the Dead Sea and the desert beyond. It is unusual among the great cities of antiquity in that very little of it was built. Most of its monuments were cut directly into the rock, and many still show the crisp edges that masons gave them two thousand years ago. For several centuries Petra was the capital of the Nabataean kingdom, an Arab state whose wealth came from the incense, spices and other luxuries carried north from southern Arabia towards the Mediterranean. The Nabataeans grew rich by taxing and protecting that trade, and their city stood where the desert routes met.

Almost every visitor enters as the traders did, through the Siq, a gorge that runs for more than a kilometre between cliffs with vertical sides. The walls above are streaked with the colours of the rock, red and purple and ochre, and the route twists so sharply that it is impossible to see far ahead. In places the passage is only a few metres wide, and the sky appears as a bright strip high above. Along one wall runs a channel that once carried water into the city. The gorge opens abruptly beneath a monumental facade about forty metres high, known as the Treasury, or Al-Khazneh, which is probably the most photographed of all Petra's monuments.

The facade is cut into the cliff rather than constructed in front of it, and the sandstone from which it was carved is still visible on either side. Despite the name, given by local people who believed that a stone urn at the top held treasure, the building was almost certainly a royal tomb. Its details combine styles familiar from Greek and Roman architecture with forms of ornament that are distinctively Nabataean, and it was probably created in the first century BC, when the kingdom was at the height of its power. A paved open space in front of it was once surrounded by other fine tombs, most of which have lost their decoration.

Beyond the Treasury lies the city proper, though little of the ordinary housing survives in a form that a visitor can recognise. The most impressive remains are tombs: Petra contains hundreds of them, from plain shafts cut into the rock to family monuments with elaborate decorated fronts. A theatre, carved from the hillside in the first century AD, held several thousand spectators, and a second monumental building, the Monastery, stands high above the valleys at the top of a long flight of steps cut into the rock.

What allowed a large population to live in such dry country was engineering. Petra has no river, and rain falls rarely, but when storms strike the hills, water pours down the valleys in sudden floods. The Nabataeans built dams to hold the floodwater back, and channels and terracotta pipes to carry it where it was wanted, along the walls of the Siq and into the city, where hundreds of cisterns and basins stored it against the dry months. Some of these structures still work today, and the scholars who study them regard the water system as being as remarkable as the tombs.

In 106 AD the last Nabataean ruler died, and the emperor Trajan annexed the kingdom, turning it into the Roman province of Arabia. Petra continued as a provincial city for a time, but its wealth had depended on trade routes that were shifting northwards. An earthquake in 363 AD brought down buildings that were never rebuilt, and over the following centuries the city shrank to a small settlement around a spring. A church with fine mosaic floors, excavated in the 1990s, shows that a Christian community lived there in the fifth and sixth centuries.

After that Petra was largely forgotten beyond its own region. In the twelfth century Crusaders built a small castle at a neighbouring site, but they did not settle in the city itself, and the place was known for the next seven hundred years mainly to the Bedouin herders who lived among the ruins and to the occasional European traveller who heard rumours of a carved city in the desert.

The visitor who brought Petra to European attention was Johann Ludwig Burckhardt, a Swiss explorer employed by a British society that promoted the exploration of Africa. Burckhardt had spent two years in Aleppo learning Arabic and the customs of the region, and he travelled in the dress of a Muslim scholar under the name Sheikh Ibrahim. In 1812, on his way from Damascus to Cairo, he persuaded a guide to take him to Wadi Musa by saying that he had promised to sacrifice a goat at the tomb of Aaron. The guide led him through the Siq, and in the shadow of the Treasury Burckhardt made notes that he kept to himself. He died in Cairo in 1817, and his account of the visit appeared in print five years later.

Petra was inscribed on the UNESCO World Heritage list in 1985, and in the same decade the Bedouin families who had lived among the monuments for generations were moved to new houses nearby. Conservation is now a constant struggle, because the sandstone absorbs water from the air and from the ground and crumbles where salts gather. Floods remain a danger, and the growing number of visitors has forced the authorities to consider how many people the narrow gorge can safely hold.

In the twelfth century a merchant who wished to buy goods in a distant city faced an awkward problem: he had to move money. Coins were heavy and bulky, and a large payment might need several pack animals. A journey of several hundred miles exposed the carrier to robbery, to extortion at toll gates and to the loss of a bag at a river crossing. Italian traders who met at the great fairs of northern France therefore developed an instrument that allowed a debt to be settled without any precious metal changing hands.

The instrument was the bill of exchange. In its classic form it is a written order by which one person, the drawer, instructs a correspondent in another city, the drawee, to pay a stated sum to a named person, the payee, on a fixed future date. Because payment is made in a different currency from the one in which the debt was incurred, the transaction does two things at once: it settles an account between two places hundreds of miles apart, and it converts one currency into another at a rate agreed in advance. The merchant buys the bill in his own city and collects the money when he, or his agent, reaches the destination.

There was a further advantage that was less openly discussed. The Church taught that lending money in return for a payment was sinful, and canon lawyers condemned the practice as usury, so that a straightforward loan at interest was forbidden to Christians. A bill of exchange, however, could yield a return that no theologian could easily identify as interest, because the profit appeared as a difference between the rates at which currencies were exchanged in two cities. A merchant who bought a bill in Venice and collected payment in Bruges might receive more than he had paid without any loan being made. Whether this was a legitimate exchange or a disguised loan was argued over for centuries.

The risks of lending on a large scale were made plain by the fate of two Florentine companies. The Bardi and the Peruzzi were among the greatest banking houses of fourteenth-century Europe, and both lent enormous sums to monarchs, above all to King Edward III of England, who was financing a long war with France. When he failed to repay his debts in the 1340s, both companies collapsed, dragging down with them many smaller investors who had entrusted money to them.

Banking on a more durable footing was the achievement of the Medici Bank in the following century. With branches in Rome, Venice, Geneva, Bruges and London, it could move funds across Europe by means of bills and settle accounts between its own offices. Its managers kept systematic records, and the technique they used, double-entry bookkeeping, was described for a wide readership in 1494 by the Franciscan mathematician Luca Pacioli, whose manual set out the method as it was practised in Venice.

For a long time a bill was payable only to the person named in it, which limited its usefulness. From the sixteenth and seventeenth centuries, however, it became common for the payee to write a signature on the back of the bill and so transfer the right of payment to somebody else. This practice, endorsement, turned a private promise into something closer to money, since a bill could pass through several hands before its due date. A holder who needed cash at once could sell the bill to a banker for slightly less than its face value, the difference being the banker's reward for waiting.

Our knowledge of these dealings comes largely from archives that were never meant to be read. The richest is the collection left by Francesco di Marco Datini, a merchant of the small Tuscan town of Prato who died in 1410: his ledgers, account books and thousands of business letters survived by accident and allow historians to follow individual transactions in remarkable detail. Notarial registers in Genoa and Venice record bills that were formally protested when payment failed. But the record is uneven, because bills that were paid were destroyed, and no clerk had a reason to preserve the ordinary business of a well-run firm.

These gaps shape the arguments among economic historians. One long-running dispute concerns how far the Church's prohibition actually restrained finance. Some scholars maintain that the ban was largely cosmetic, since a merchant could always find a device to disguise a loan; others reply that the doctrine genuinely limited what respectable firms would do and pushed credit into particular forms. A second debate concerns whether such instruments caused the expansion of trade or merely followed it. Similar devices existed in other trading cultures, notably in the Islamic world, and historians disagree about whether these were independent inventions or connected through long-distance commerce.

In 1609 the city of Amsterdam founded an exchange bank to which merchants brought their varied coins and received credit in a single stable unit of account, with the result that a bill drawn on Amsterdam was worth more than one drawn almost anywhere else.

By the nineteenth century the centre of the trade had moved to London, where specialist discount houses bought and sold bills on a vast scale. The bill on London became the ordinary currency of international commerce, accepted in distant markets because it could always be sold in the city where it was drawn.

In 2023 a competition was launched with a simple-sounding aim: to read a book that no one could open. The object in question was one of the papyrus scrolls found in the eighteenth century at Herculaneum, a Roman town buried by the eruption of Vesuvius in 79 AD. Heat carbonised the scrolls, and they are now so brittle that any attempt to unroll them reduces them to fragments. The competition offered large prizes to any team that could recover a passage of Greek text from a scroll without touching it. Within a year, groups using machine learning had done so, and hundreds of words had been read from a scroll that remains, physically, closed.

The library that produced the scrolls is thought to have belonged to a luxurious villa on the edge of the town. Excavators in the 1750s recovered around 1,800 scrolls, most of them works of Greek philosophy, and early attempts to open them destroyed many. For two and a half centuries scholars could do little more than guess at the contents of the rest from the few that had been successfully unrolled, often at terrible cost. What makes the problem so hard is the ink, which is essentially carbon, the same element as the papyrus itself, so that an X-ray image shows no difference between the letters and the material beneath them.

The new method does not read ink directly. A scroll is first scanned in a synchrotron, a machine that produces extremely intense X-rays, and the scan records minute variations in density produced by the layers of tightly rolled papyrus. Software then works out how the sheet was wrapped and flattens it, layer by layer, into a virtual page. Finally a machine-learning model, trained on images of papyri whose ink is known, searches for the faint patterns that the letters leave on the surface. Similar techniques had earlier succeeded with a scroll from En-Gedi, near the Dead Sea, but there the ink contained metal, which X-rays can see; the Herculaneum ink offers no such help.

Texts on stone pose a comparable problem. Greek inscriptions often survive with letters, and sometimes whole words, missing where the surface has broken away, and editors have traditionally filled the gaps with suggestions based on grammar, on parallel passages and on a sense of what a stonecutter would have written. A system trained on tens of thousands of inscriptions has been built to make such suggestions automatically. Its designers reported that it proposed the correct text in a majority of cases, and that it performed better still when a historian chose among the options it offered.

The case for using these tools is straightforward. The alternative to a machine-assisted restoration is usually not a certain text but silence. If a model can suggest plausible words to fill a gap, a scholar gains a hypothesis to test, and a hypothesis can be checked against metre, grammar and other inscriptions. The practice also has an honourable precedent: editors have always supplied lost letters, and they mark their supplements with brackets to show that they are not part of the stone or the papyrus itself.

Yet there is a difficulty at the heart of the method. A model learns from the texts that have survived, and it predicts what is most likely given what it has seen. The passages historians most want to recover are often precisely those that are least like anything else: an unusual argument, a rare word, a line that contradicts what we thought an author believed. On such passages the model is at its weakest, and its errors are least visible, because a smooth and conventional restoration reads perfectly well.

A second danger is that a suggestion will be mistaken for a reading. A reconstruction quoted without its brackets, or a machine's preferred word repeated in a popular account, can harden into a fact within a few years. The risk is greatest where money is involved, because the market in antiquities has long rewarded anyone who can produce a text, and a convincing forgery is far easier to manufacture with a model's help than with a chisel. Both tendencies argue for strict conventions of disclosure.

There is a further, quieter concern. Spectacular technical projects attract publicity and funding, while the patient work of editing, collating manuscripts and cataloguing fragments attracts neither. A discipline that rewards the announcement of a deciphered word more than the long labour of checking it may find that its most valuable skills are neglected, and it is those skills that decide whether a restoration is right.

My own view is that the tools should be used, and used boldly, but that the boundary between reading and guessing must be kept visible. A restoration proposed with the help of a model should be published together with the model's confidence and the alternatives it considered, and it should never be cited as evidence of what an ancient author actually wrote. The machine can suggest; only a scholar, accountable to the evidence, can decide. That division of labour is not a limitation of the new methods but the condition of their usefulness.