Short-answer questions

Reading Passage 2 – Second Lives Beneath the Waves

A When a ship reaches the end of its working life, it is usually towed to a breaker's yard, where it is cut up and its steel sold for scrap. A small number, however, meet a very different fate. Cleaned, stripped and then deliberately sunk, they come to rest on the seabed, where within months they begin to attract fish, corals and sponges. Artificial reefs of this kind have been created off the coasts of the United States, Australia, Malta and many other countries, and they have become popular destinations for divers. But the practice has always had its critics, and the question of whether a sunken ship is a gift to the ocean or simply a convenient way to dispose of waste has never been entirely settled.

B The idea of placing objects in the sea to attract fish is far from new. Fishermen in Japan were sinking wooden frames weighted with stones in the eighteenth century, and similar practices have been recorded in the Mediterranean. What has changed is the scale. In the second half of the twentieth century, governments began to sink not only ships but also old railway carriages, military vehicles and even concrete from demolished buildings. Some of these early projects were poorly planned: in one notorious case off the coast of Florida, around two million used car tyres were bound together and dropped into the sea in the 1970s, in the hope that they would become a reef. Instead, the bundles broke apart in storms, and loose tyres were thrown against existing natural reefs, damaging them. Clearing them up has cost millions of dollars.

C Lessons from such failures have shaped modern practice. Before a ship can be sunk today, it must be thoroughly cleaned of fuel, oil, paint containing toxic substances, and any loose material that could float away. Wiring and insulation, which may contain harmful chemicals, are removed, and openings are cut in the hull so that divers can enter and leave safely. This preparation is expensive and can take more than a year; in some cases it costs more than the value of the scrap metal the ship would otherwise have produced.

D Why do marine organisms colonise a wreck so quickly? Part of the answer is that, on much of the seabed, hard surfaces are scarce. Many coastal waters have floors of sand or mud, to which corals and sponges cannot attach themselves. A ship provides a large, stable structure rising from the seabed, with sheltered spaces of every size, from narrow gaps between metal plates to entire rooms. Small fish shelter in these spaces, and larger predators follow them. Biologist Elena Marsh, who has surveyed wrecks in the Gulf of Mexico for over a decade, has found that the number of fish around a ship can exceed that on a natural reef of similar size within five years.

E Not everyone is convinced that this represents a real benefit. A central question, raised by the fisheries scientist Tom Akers among others, is whether artificial reefs actually produce new fish or merely attract existing ones from the surrounding area. If the latter is true, a wreck may do more harm than good, because concentrating fish in one place makes them easier for fishing boats to catch. Akers argues that the answer probably varies from one species to another, and that without careful long-term monitoring it is impossible to know which effect dominates in any particular case.

F There are also concerns about what happens as a wreck ages. Steel corrodes in seawater, and while a large ship may remain structurally sound for many decades, its upper decks will eventually collapse. The engineer Rosa Lindqvist points out that the rate of corrosion depends heavily on depth and water temperature, and that ships sunk in shallow, warm waters may deteriorate much faster than their planners expected. As the structure breaks down, the habitat it provides changes too, and divers who enter older wrecks face increasing risks.

G Supporters respond that, whatever its effect on fish numbers, a well-managed artificial reef can relieve pressure on natural ones. Popular natural reefs suffer damage from the anchors and fins of large numbers of visiting divers. By offering an attractive alternative nearby, a sunken ship can draw some of these visitors away. The marine economist David Okoro has calculated that a single wreck off the coast of South Australia generates several million dollars a year in tourism, much of it for small businesses such as dive schools and guesthouses. For coastal communities, he suggests, this income may matter as much as any ecological argument.

24. What did fishermen in Japan sink in the eighteenth century to attract fish? (24) ______

25. What items were dropped into the sea off Florida in the 1970s? (25) ______

26. Which part of a sunken ship will eventually collapse? (26) ______

Reading Passage 3 – In Defence of Forgetting

A It has become something of a commonplace to treat forgetting as the regrettable price we pay for having a memory at all: an unavoidable leakage, to be minimised by mnemonic techniques, dietary supplements or, increasingly, by delegating recollection to our devices. The assumption underlying this view – that the ideal memory would be a complete and permanent record – is seldom examined. Yet a substantial and still expanding body of work in cognitive psychology and neuroscience implies that it may be fundamentally misconceived. On this alternative account, forgetting is not merely what happens when the machinery of memory falters; it is, in a great many cases, an active and adaptive process, without which remembering itself would be considerably less useful.

B The empirical study of forgetting is conventionally dated to the 1880s, when Hermann Ebbinghaus, using himself as his sole subject, committed to memory long lists of nonsense syllables and measured how much he could relearn after intervals ranging from twenty minutes to a month. The resulting 'forgetting curve' – steep at first, then flattening progressively – has been replicated often enough to be regarded as one of psychology's more robust findings. It is customarily cited as a measure of the brain's fallibility. That reading is not wrong, exactly, but it is partial. A curve of this shape is equally consistent with a system that provisionally retains incoming information for long enough to establish whether it recurs, and then discards most of what does not. Seen in this light, rapid early loss looks less like a defect than like a sensible allocation of finite resources.

C Recent neurobiological research lends some weight to this interpretation. Experiments on animals have identified molecular mechanisms that appear to weaken the connections between neurons actively rather than simply allowing them to decay, and in some studies, interfering with these mechanisms has made it harder, not easier, for animals to adjust when the location of a reward is changed. Caution is warranted in extrapolating from rodents to humans, and the precise significance of these findings remains contested. Nevertheless, they make it difficult to sustain the assumption that forgetting is invariably passive.

D The functional advantages of losing information are perhaps clearest in the formation of concepts. A child who encounters dozens of individual dogs does not retain an exhaustive file on each; instead, the idiosyncrasies of particular animals fade while the regularities they share are consolidated. What emerges is an abstraction flexible enough to accommodate a breed the child has never seen. The point is not trivial: a mind that registered every dog as an entirely separate case would be poorly equipped to cope with novelty, which is, after all, the condition under which most decisions have to be made. Some theorists go further and argue that the erosion of detail is not a by-product of generalisation but a precondition for it. Researchers in machine learning have arrived, by a quite different route, at a strikingly similar conclusion: a model that reproduces its training data too faithfully – a problem known as 'overfitting' – tends to perform poorly on cases it has not previously encountered.

E A rather different argument concerns the emotional colouring of memory. There is reasonable evidence that the affective intensity of unpleasant recollections tends to diminish more rapidly than that of pleasant ones, a pattern that may help to explain why most people are able to recover from humiliation or loss. It would be simplistic to attribute such recovery to forgetting alone; social support, reinterpretation and the mere passage of time all play their part, and the relative contribution of each is far from settled. Still, it is hard to resist the conclusion that the gradual softening of painful detail is, for the majority, more a mercy than a malfunction.

F None of this should be taken as a blanket endorsement of forgetting. Neurodegenerative conditions such as dementia inflict profound suffering, and it would be perverse to suggest that a student who cannot recall material on the day of an examination has been well served by her brain. Yet even in education, an appreciation of forgetting has practical consequences. Ebbinghaus himself noticed that material reviewed across several sessions was retained far better than material crammed into one, a phenomenon now known as the 'spacing effect'. Paradoxically, the benefit appears to be greatest when some forgetting has been allowed to occur between sessions: the effort of retrieving partially faded information seems to strengthen it more than effortless rereading does. Forgetting, in other words, is not only compatible with durable learning but may be one of its ingredients.

G What, then, follows? Certainly not that we should cease trying to remember, nor that the digital tools on which we increasingly depend are harmful; they are simply poor models of the faculty they supplement. The more modest conclusion is that the prevailing metaphor of memory as a storage device – a metaphor reinforced by every hard drive and cloud archive we use – obscures more than it illuminates. A file is either retained or deleted; biological memory, by contrast, appears to be engaged in a continuous process of evaluation, strengthening some traces and allowing others to recede in the light of what experience suggests will prove useful. To regard each lapse as evidence of deficiency is to misunderstand what memory is for.

39. What term describes the problem of a computer model copying its training examples too closely? (39) ______

40. According to the writer, what kind of image of memory gives a misleading picture of how it works? (40) ______