IELTS Academic Reading Practice Test 472

On 21 December 1913, readers of the New York World opened their copy of Fun, the supplement that came with the Sunday edition, to find a diamond-shaped grid of empty squares. The instructions invited them to fill the grid with words running across and down, using the numbered clues printed beneath it. The puzzle had been devised by Arthur Wynne, a journalist born in Liverpool who had worked for newspapers in Britain before moving to the United States. Wynne had been asked to fill a space in the Christmas number of the supplement. His grid had no black squares, now a standard feature of the puzzle, and its squares were numbered to match the clues.

The idea was not entirely new. Victorian magazines had printed word squares, in which the same words read across and down, and variants of such puzzles appeared in American children's periodicals in the nineteenth century. Wynne's contribution was the interlocking grid: every letter had to work in two directions at once. Numbering the squares, rather than blacking out the unused ones, meant that the solver worked from a list of clues instead of relying on guesswork.

Readers of the World responded with enthusiasm, and before long they were sending in puzzles of their own, which the paper printed. The game's name was an accident. It had been billed as a 'word-cross', but a typesetter apparently reversed the two halves, and 'crossword' appeared in print. The error was never corrected, and the new name stuck. Within a few years the crossword had become a regular feature of the paper, and imitators had appeared in rival publications.

In 1924 two young publishers, Richard Simon and Max Schuster, were approached by a relative of Simon's who wanted a collection of the puzzles in book form. The partners were doubtful: they could not see why anyone would buy in a book what could be read in a newspaper. They printed a modest first run, attaching a pencil to each copy in the hope of making the purchase irresistible. The book sold out almost at once and was followed by others. According to the firm, the first title had sold several hundred thousand copies by the end of the year, and collections of puzzles dominated the bestseller lists.

What followed was a craze. In the United States, commuters were said to be missing their stops, and office managers complained that staff were solving puzzles at their desks. Some commentators treated the fashion as a harmless amusement; others were less tolerant, writing to newspapers to complain that the puzzle wasted time and distracted the young from their studies. The fashion crossed the Atlantic quickly, and British papers soon began printing grids of their own.

The New York Times was the most conspicuous holdout. In 1924 it published an editorial dismissing crosswords as a trivial and faintly shameful addiction, and for years afterwards it refused to print one. The attack did not last. After the Japanese attack on Pearl Harbor in December 1941, the appetite for distraction grew, and in February 1942 the paper's Sunday magazine carried its first crossword, edited by Margaret Farrar, who as Margaret Petherbridge had made her name as an editor of puzzle books. It became a weekly fixture, and it is now widely regarded as the most prestigious puzzle of its kind.

In Britain the puzzle took a different form. The Times carried its first crossword in 1930, and the early grids were straight, literal puzzles of the American kind. But British setters, encouraged by a readership that enjoyed verbal games, began to devise clues that worked on two levels: a definition of the answer combined with some form of wordplay — an anagram, a hidden word, or a pun on how a word sounds. This cryptic style became a British speciality.

The most influential of the early setters was Edward Powys Mathers, who wrote puzzles for The Observer under the pen name Torquemada and who is credited with establishing many of the conventions of the cryptic clue. His puzzles were famously difficult. Later setters codified the rules: each clue must contain a definition at one end or the other, and the remaining words must supply the letters through a recognised device.

Crosswords have also been put to less innocent uses. During the Second World War, British intelligence officers are said to have sent prisoners of war in Germany puzzles constructed to carry hidden messages, slipped into parcels delivered to the camps. The scheme worked because a completed grid of letters can conceal information that looks like nothing more than the answer to an obscure clue.

Today the crossword is a global industry. In the United States, where grids are still largely symmetrical and clues usually literal, constructors speak of 'crosswordese' — the small stock of short, vowel-rich words, many of them of foreign origin, that can be slipped into awkward corners when no ordinary word will fit. In Britain the cryptic puzzle dominates. The two traditions have developed separate conventions, separate competitions and even separate vocabularies, and the American Crossword Puzzle Tournament, founded in 1978, now attracts several hundred solvers each year.

More than a century after Wynne's grid appeared, the appeal of the crossword seems to lie in its combination of order and surprise: a small, closed world in which every square has exactly one right answer.

A landslide is the downslope movement of rock, soil or debris under the influence of gravity. The term covers an extraordinary range of events. At one end are the slow creeps that shift a hillside by a few millimetres a year, cracking walls and distorting roads over decades; at the other are rock avalanches that release millions of cubic metres of material and travel several kilometres in a few minutes. Between them lie slumps, in which a block of ground rotates forward on a curved surface, debris flows that resemble a river of wet concrete, and rockfalls that begin with the detachment of a single block from a cliff face. Engineers classify such events by the material involved and by the way it moves, because each type behaves differently.

Whether a slope stays where it is depends on a contest between two sets of forces. Gravity pulls the material downwards, and the steeper the slope, the larger the share of that pull acting along it. Against this driving force the ground mobilises resistance: friction between grains, and cohesion, the tendency of particles to stick together, which is supplied in soil by clay minerals and by the roots of plants. Engineers describe the ratio of resisting to driving forces as the factor of safety; when it falls below one, movement begins. Dry sand will not hold at an angle steeper than about thirty to thirty-five degrees, its angle of repose, and finer or rounder grains hold even less. Add a little moisture, however, and the same sand will stand much more steeply, because surface tension between the grains briefly acts like glue.

Water is the most common cause of failure, and it works in several ways at once. Rain adds weight to the material, increasing the driving force. More importantly, it raises the pressure of water in the pores between grains. That pressure pushes the grains apart, reducing the contact between them and therefore the friction that holds the slope together. A soil that was strong when damp may lose much of its strength once it is saturated, and the change can be abrupt: a slope that has survived many ordinary storms may give way during a single exceptionally wet month. The most dramatic case is quick clay, which forms in marine deposits later lifted above sea level. Rainwater has gradually washed the salt out of it, leaving a structure that is stable until it is disturbed, at which point it can collapse into a liquid that flows almost like water.

Not every slope fails because of rain. Earthquakes shake slopes and raise pore pressures so rapidly that saturated sandy ground can liquefy, and many of the landslides that follow a large earthquake begin within seconds of the shaking. Volcanoes destabilise their own flanks, and the collapse of a volcanic edifice can send a debris avalanche across a plain far below. Rivers and waves steadily remove material from the base of a slope, steepening it until it can no longer stand, and a single flood can do in an afternoon what ordinary erosion would take centuries to achieve. Human activity often accelerates the same processes. Cutting a road into a hillside removes support from below; building at the top adds weight; clearing trees takes away the roots that bind the soil; and a leaking pipe can saturate ground that would otherwise stay dry.

Where large populations live beneath steep ground, the consequences can be appalling. In October 1963 a block of rock and soil several hundred metres wide slid into the reservoir behind the Vaiont dam in northern Italy, displacing so much water that waves overtopped the dam and swept through the villages below; almost two thousand people died. The dam itself held, which is why the disaster is remembered as a failure of warning rather than of engineering: the slope had been moving for weeks, and the danger had been recognised without being acted upon. A quick-clay landslide at Rissa in Norway in 1978, in which a photographer happened to be filming, showed how suddenly an apparently ordinary hillside can turn into a flow.

Because failure is usually preceded by movement, the most reliable warning comes from measuring the ground itself. On the surface, surveyors use satellite positioning to detect shifts of a few millimetres, while radar instruments that bounce signals off a hillside can map movement across an entire valley from a safe distance. Satellites carrying synthetic aperture radar now image the same slope every few days, and a technique called interferometry compares successive images to reveal small changes in height. Below ground, instruments lowered into boreholes record the tilt of the rock, the widening of cracks and the pressure of water in the pores. None of this is cheap, and it is usually deployed only where a slope has already been identified as a threat. In Hong Kong, where steep slopes lie close to dense housing, a warning system introduced after a disaster in 1972 combines rainfall measurements with a network of instruments and a programme of repairs.

Yet prediction remains inexact. Two slopes of identical shape and composition may behave quite differently, because the strength of the ground depends on details — cracks, old slip surfaces, the fabric of the clay — that can never be mapped completely. Movement is not a dependable sign either: some slopes creep for years without failing, while others collapse after no measurable warning at all.

In most examinations a mark is supposed to mean something in itself: a student who scores seventy per cent has shown a certain amount of knowledge, whether the rest of the class did better or worse. Under a different arrangement, known as norm-referenced grading or grading on a curve, marks are adjusted so that a fixed proportion of students receive each grade. The best tenth might be given the highest grade, the next twenty per cent the next, and so on, whatever marks they actually achieved. The practice is used in some university departments and law schools, and a version of it has been applied by employers ranking their staff.

Defenders of the practice make three arguments. The first is that a curve protects the value of a qualification. If a paper turns out to be unusually easy, raw marks rise, and the top grade stops distinguishing an excellent student from a merely competent one; scaling the results restores the distinction. The second is that a curve acts as quality control on the examination itself, compensating for a paper that was badly designed. The third concerns scarcity: if a university can admit two hundred students and three thousand apply, somebody must decide who ranks higher, and comparison within the applicant group is, on this view, the least bad method available.

Specialists in educational measurement are less confident. Every test produces a score with a margin of error: a student who scores 68 on one occasion might reasonably have scored 64 or 72, depending on which questions came up and how generously a marker read a half-made point. Near the middle of a distribution these uncertainties matter little, but at the boundaries they matter a great deal. If a grade boundary falls inside that margin, the difference between an A and a B may reflect luck rather than competence. A curve that imposes a fixed boundary turns this noise into a permanent judgment, and does so without anyone being able to show that the judgment is wrong.

Critics concerned with fairness raise a different objection. Under a curve, a student's grade depends not only on their own performance but on everyone else's. A capable student in an exceptionally strong year group can be placed below a weaker student in a weak one, and neither has any control over the company they keep. The arrangement punishes a student for the accident of who else joined the course, and it rewards departments that admit weaker intakes, which is the reverse of what most people expect a grading system to do. Grades, these critics argue, stop being a statement about what a student can do.

Employers and admissions tutors are caught in the middle. What they want is a signal they can trust. Presented with a transcript in which seventy per cent of applicants hold the top grade, a selection panel learns very little, and it is hardly surprising that some firms have responded by setting their own tests or by favouring universities whose standards they believe they can gauge. Yet a curve does not solve their problem either: a grade awarded in a strong cohort may represent more ability than a higher grade awarded in a weak one, and the employer has no way of telling which is which.

Professional licensing bodies have avoided the difficulty by separating two questions. The first is whether a candidate is competent: can this person safely be trusted to prescribe a drug or to certify the strength of a bridge? Here a pass mark is fixed in advance, often by asking experienced practitioners to estimate how a borderline candidate would perform on each question, and everyone who clears it passes. The second question is who should be appointed to a particular post, and there ranking cannot be avoided.

The lesson, it seems to me, is that the case for a curve depends on the purpose of the assessment, and that the error lies in applying one model to the other's task. Where the purpose is certification — has this student reached the standard? — a fixed distribution is indefensible, because it can fail a candidate who has met the standard and pass one who has not. Where the purpose is selection, comparison is necessary, and the honest course is to say so and to publish the evidence on which it rests.

None of this means that the defenders' worries can be dismissed. Grade inflation is real. In the United Kingdom the proportion of first-class degrees awarded roughly doubled during the 2010s, and few people believe that students became twice as able. A system in which grades rise without limit eventually tells employers and the public nothing at all. But the remedy is not a compulsory curve, which suppresses the symptom while leaving the underlying problem — the absence of agreed standards — untouched. The remedy is external scrutiny of marking, and transparency about results.

A defensible system would award marks against published criteria and require examiners to show that a script meets them. Alongside each student's mark, institutions should publish the distribution of marks for the cohort, so that a reader can see where that mark sits without the mark itself being altered. Comparison then remains possible, but it no longer decides who passes.

The curve, in short, is a device for ranking people that has been mistaken for a device for judging them. It should be abolished not because competition is distasteful, but because it answers a question that examinations are never asked.