Learning to Sing: How Young Birds Acquire Their Songs
A male chaffinch raised in isolation, hearing no other bird of its kind, will still produce a song when it matures — but a strange, impoverished one, lacking the crisp flourishes that mark the species in the wild. Reared instead within earshot of adult chaffinches, the same bird grows up to sing normally. This simple contrast, established by experiments in the mid-twentieth century, overturned a comfortable assumption. Birdsong, it turned out, is not simply wired into the animal at hatching. For a large group of birds — the songbirds, or oscines, which include finches, sparrows and thrushes — the full, characteristic song must be acquired by listening to others, much as a child acquires speech. The discovery opened one of the most productive lines of enquiry in the study of animal behaviour.
That acquisition is not open-ended. In many species there is a limited stretch early in life, often within the first few months, during which the young bird must be exposed to adult song if it is ever to sing properly. Researchers call this the sensitive period. A bird that hears the correct model within this window commits it to memory and will later reproduce it; a bird that hears nothing until the window has closed cannot make up the loss, however much song it is played afterwards. The memory formed during this phase serves as a template — an internal standard against which the bird will subsequently measure its own efforts. Different species vary widely in how rigid this timetable is; some remain open to new songs well into adulthood.
Crucially, storing the template and producing the song are separated by weeks or even months. A young bird does not burst into accurate performance. Instead it begins with what is called subsong, a rambling, quiet and formless twitter that bears only a faint resemblance to the adult version. Over time this raw material is gradually shaped, passing through a more structured 'plastic' stage before settling, or crystallising, into the fixed adult song. The parallel with a human infant's babbling, which slowly firms up into words, is one that scientists have drawn repeatedly, and it is more than a loose metaphor. Quite why the process should be so protracted is not fully understood, but the delay appears to give the bird room to experiment.
For this shaping to work, the bird must be able to hear itself. If a young songbird is surgically deafened after it has memorised the template but before it has begun to sing, its output remains a jumble; it never converges on the model it once stored. The bird, in other words, is listening to its own attempts and correcting them against the remembered standard, closing the gap by trial and error. Once the adult song has crystallised, however, this dependence relaxes: a bird deafened at that later stage will, for a time, continue to sing its established song more or less intact.
Because song is learned rather than inherited whole, it can drift and diverge from place to place, exactly as spoken language does. Populations of a single species living in different valleys frequently sing recognisably different versions — local dialects that a trained ear, or a sound spectrograph, can tell apart. A young male typically adopts the dialect of the neighbourhood in which he settles, not necessarily the one his father sang. These regional variants can persist across generations, handed down by imitation, and they occasionally mark boundaries that females appear to notice when choosing a mate. In a few well-studied species, the number of distinct dialects across a region runs into the dozens.
The appeal of songbirds to scientists goes well beyond the birds themselves. Very few animals learn the sounds they make by imitation; most, including our closest primate relatives, are born with a fixed repertoire of calls. Vocal learning of the kind songbirds display is otherwise found in only a scattering of groups — among them parrots, hummingbirds, whales and human beings. Because a songbird's learning unfolds on a convenient timescale and in a brain small enough to study in detail, it has become a favoured model for asking how any nervous system, ours included, turns what it hears into what it can produce. What is learned in a matter of weeks by a bird may, studied closely enough, illuminate what takes a human child years.
Reading Passage 1 has six paragraphs, A–F. Choose the correct heading for each paragraph from the list of headings below. Write the correct number, i–ix, next to each paragraph.
- i A limited window for learning
- ii How most other animals inherit their calls
- iii Evidence that the song is not inborn
- iv The role of feedback from one's own singing
- v From formless practice to a finished song
- vi Regional variations in a single species' song
- vii Why songbirds matter to science beyond themselves
- viii The physical structure of the bird's vocal organ
- ix Differences between male and female singers
Do the following statements agree with the information given in Reading Passage 1? Write TRUE if the statement agrees with the information, FALSE if it contradicts it, or NOT GIVEN if there is no information on this.
Complete the sentences below. Choose NO MORE THAN TWO WORDS from the passage for each answer.
Saving for Tomorrow: The Psychology of Putting Money Aside
Ask people whether they ought to save more for the future and most will say yes. Ask whether they actually do, and the answer is frequently no. Surveys across wealthy countries repeatedly find households with too little put aside for retirement or for an emergency, even among those who earn comfortably and insist that saving matters to them. Classical economic theory struggles with this. It assumes a rational individual who weighs present and future needs and settles on the level of saving that serves them best over a lifetime. If that were an accurate picture, the persistent gap between what people intend to save and what they manage to save would not exist. Behavioural economics grew, in part, out of the effort to explain it.
A central culprit, in this account, is the way human beings value time. Offered a choice between a smaller reward now and a larger one later, people tend to prefer the immediate payoff far more strongly than a purely rational calculation would justify. Economists call this present bias. Crucially, the preference reverses depending on distance: asked in advance to choose between a treat in a year and a bigger treat in a year and a week, most people happily wait the extra week; but when the year has passed and the smaller treat is available today, the same people grab it. Saving asks us to hand a benefit to a future self we find oddly difficult to care about, while the costs fall on the self of the present.
This tug-of-war is often described as a conflict between two versions of the same person: a far-sighted planner who resolves, each January, to economise, and an impulsive doer who, each afternoon, spends. The planner is real, but weak; the doer generally wins. What follows from this framing is important. If the problem were simply ignorance — people not knowing that they should save — then information would fix it. But those who overspend usually know perfectly well what they should be doing. The obstacle is not a lack of knowledge but a lack of self-command, and remedies aimed at educating people therefore tend to disappoint.
If willpower is scarce, one solution is to design systems that require less of it. The most celebrated example concerns workplace pensions. When employees have to opt in — to fill in a form and actively choose to join a savings scheme — participation is often strikingly low, not because workers object but because the task sits perpetually at the bottom of a to-do list. Switch the arrangement so that employees are enrolled automatically and must opt out if they wish to leave, and participation leaps, frequently above ninety per cent. The same human inertia that kept people out of the scheme now keeps them in. Nothing has been forbidden and no one has been paid; only the default has changed.
Other interventions work with present bias rather than against it. One influential scheme invites employees to commit, today, to saving not from their current pay but from future pay rises they have not yet received. Because the sacrifice lies in the future, the present-biased self does not resist it; and because the extra saving is taken from money the worker never had in hand, it is not experienced as a loss. Programmes built on this principle have raised saving rates substantially. They exploit a quirk of the mind — that money feels different depending on how it is labelled and when it arrives — turning what is usually an obstacle into a lever.
Not everyone is persuaded that psychology is the heart of the matter. A more sceptical camp argues that the emphasis on mental quirks risks blaming individuals for what are, at bottom, structural problems. Many households do not save, on this view, for the simple reason that after rent, food and bills there is nothing left to save; no clever redesign of a pension form will conjure a surplus that does not exist. There is force in this objection. Yet the two explanations need not compete. Even where incomes are adequate, the behavioural obstacles are real, and even where the obstacles are removed, low incomes will still bite. The insights of behavioural economics are a supplement to policies that raise incomes, not a substitute for them.
What the field has established, beyond much dispute, is that how a choice is presented can matter as much as the choice itself. Two schemes offering identical financial terms can produce wholly different results depending on which option is the default, when the money is taken and how it is described. That lesson has spread far beyond pensions, into health, energy use and public policy of many kinds. Its critics warn that nudging people towards decisions someone else has judged good for them raises questions about autonomy that are not easily dismissed. The debate over where helpful design ends and manipulation begins is likely to outlast the novelty of the idea.
Choose the correct letter, A, B, C or D.
Look at the following statements and the list of groups below. Match each statement with the correct group, A–D. NB You may use any letter more than once.
- A Classical economic theory
- B Behavioural economists
- C Sceptics of the behavioural approach
- D Critics of nudging
Complete the summary below. Choose NO MORE THAN TWO WORDS from the passage for each answer.
Do the following statements agree with the claims of the writer in Reading Passage 2? Write YES if the statement agrees with the claims of the writer, NO if it contradicts them, or NOT GIVEN if it is impossible to say what the writer thinks.
The Clockwork That Cannot Be Read: Chaos and the Limits of Prediction
For three centuries after Newton, the success of physics fed a powerful conviction: that the universe, at bottom, is a clockwork mechanism whose future is fixed by its present. The most famous statement of this faith belongs to the French mathematician Pierre-Simon Laplace, who imagined an intellect vast enough to know the position and motion of every particle. To such an intellect, he wrote, nothing would be uncertain; the future, like the past, would lie open before its eyes. Prediction, on this view, was limited only by ignorance. Gather enough data, refine the equations, and in principle any system could be forecast as reliably as an eclipse. Chaos theory, which took shape in the second half of the twentieth century, did not overturn the determinism behind this picture. What it overturned was the assumption that determinism guarantees predictability.
The turning point came, famously, by accident. In the early 1960s the meteorologist Edward Lorenz was running a simplified computer model of the atmosphere. Wishing to repeat a portion of a calculation, he re-entered a set of numbers from a printout, rounding one value from six decimal places to three — a difference he assumed too trivial to matter. The weather his model then produced diverged, within a simulated month or two, into something utterly unlike the original run. A rounding error smaller than a thousandth had been magnified into a completely different forecast. Lorenz had stumbled on what is now called sensitive dependence on initial conditions: in certain systems, differences too small to measure grow, relentlessly, until they dominate the outcome. The popular image of a butterfly's wingbeat eventually shifting the path of a distant storm captures the idea exactly.
It is tempting to conclude that such systems are simply random, but this is precisely the wrong lesson, and the confusion is worth dispelling. A chaotic system is entirely deterministic: it obeys fixed rules with no element of chance, and the same starting point will always yield the same result. Randomness plays no part. The trouble is that we can never specify the starting point with perfect precision — every measurement carries some error, however small — and in a chaotic system that unavoidable smudge of uncertainty is amplified so rapidly that long-range prediction becomes impossible in practice. The system is not unpredictable because it is disorderly; it is unpredictable because it is exquisitely, and deterministically, sensitive. This distinction, subtle as it sounds, lies at the heart of the subject.
The consequence is that each chaotic system carries what might be called a prediction horizon — a stretch of time beyond which forecasting decays into guesswork, no matter how good the model or the measurements. For the atmosphere, this horizon is thought to be around two weeks. It is not that meteorologists lack skill or computing power; the limit is a property of the system itself, and no conceivable improvement in instruments will push a detailed daily forecast much past that boundary. Doubling the accuracy of our measurements does not double the range of reliable prediction — it extends it only slightly, because the errors grow so fast. This is why a forecast for tomorrow can be dependable while one for the same date next month is worthless.
Yet it would be a mistake to picture chaos as mere formlessness, and here a second, more hopeful side of the theory emerges. When the behaviour of a chaotic system is traced not moment by moment but as a whole, over long spans, it often reveals a hidden order: its states, plotted in the abstract, trace out intricate but bounded shapes that mathematicians call strange attractors. The individual path is unpredictable; the overall pattern is not. A climate, for instance, may be describable in its broad statistics — its typical range, its averages — even when tomorrow's weather within that climate cannot be pinned down. Some scientists therefore stress that chaos does not abolish order so much as relocate it, moving it from the level of the single event to the level of the long-run pattern. Prediction, on this reading, is not defeated but redefined.
Once researchers knew what to look for, chaotic behaviour turned up nearly everywhere: in the beating of the heart, in the swings of animal populations, in fluid turbulence, in the wobble of certain planetary orbits over vast timescales. This ubiquity carries a sobering lesson for anyone who forecasts. Many of the systems whose futures we most wish to know — economies, ecosystems, the climate over centuries — contain chaotic elements, which places firm limits on how far ahead detailed prediction can reach. But the same body of work distinguishes sharply between two questions that are easily muddled: what a system will do next, and what it will tend to do on average. The first may be forever beyond us; the second may be entirely within reach. Confusing the two produces both false confidence and needless despair.
The deepest legacy of chaos theory is thus not a set of equations but a correction to an old ambition. Laplace's demon fails, it turns out, not because the world is undetermined but because knowledge of the present can never be complete, and in a chaotic system incompleteness, however slight, is fatal to long-range foresight. Determinism and predictability, so long assumed to be the same thing, come apart. A universe can be rigidly law-governed and yet, for any observer trapped inside it, largely unforeseeable. Whether this counts as a defeat for science or a maturing of it is a matter of temperament. What is no longer in doubt is that the confident forecasting dreamed of in Laplace's age was never, even in principle, available to beings like us.
Reading Passage 3 has seven paragraphs, A–G. Which paragraph contains the following information? Write the correct letter, A–G. NB You may use any letter more than once.
- A Paragraph A
- B Paragraph B
- C Paragraph C
- D Paragraph D
- E Paragraph E
- F Paragraph F
- G Paragraph G
Do the following statements agree with the information given in Reading Passage 3? Write TRUE, FALSE or NOT GIVEN.
Choose the correct letter, A, B, C or D.
Answer the questions below. Choose NO MORE THAN THREE WORDS from the passage for each answer.