Academic Reading 1: night trains, concrete, and the practice debate

Passage 1 · The return of the night train

For most of the twentieth century the sleeper train was an ordinary way to cross a continent. A traveller boarded in one city after dinner, slept while the country went past, and stepped onto a platform several hundred kilometres away in time for a morning meeting. By the late 1990s the service was disappearing almost everywhere in Europe. Budget airlines had made the same journey cheaper and faster, and the carriages themselves — heavy, specialised and used for only one trip a night — were expensive for an operator to own.

The decline was rapid. Between 2001 and 2016 the number of night services in Europe fell by roughly two thirds, and several countries abandoned them altogether. Germany's national operator withdrew from the market in 2016, having concluded that the business could not be made to pay. Sleeping cars were sold, scrapped, or left in sidings, and the skills needed to maintain them left the industry with them.

What happened next surprised almost everybody. Austria's national railway bought a quantity of the German rolling stock and began running the routes its neighbour had given up. The Austrian operator was not sentimental about it: its calculation was that a night train fills a bed rather than a seat, and that a passenger who sleeps on board does not pay for a hotel. Priced against a flight plus a night's accommodation rather than against a flight alone, the sleeper looked considerably better. By 2019 the Austrian network was carrying well over a million passengers a year on services that another operator had judged unviable.

Two further changes have worked in the sleeper's favour, and it is worth separating them. The first is environmental. A long-distance rail journey typically produces a small fraction of the emissions of the equivalent flight, and in several northern European countries a measurable number of travellers now say they avoid short-haul aviation for that reason. The second is simple frustration with the airport. A flight that takes ninety minutes in the air routinely takes five hours door to door once the journey to the airport, the queue at security and the wait for luggage are counted. A sleeper that leaves a city centre at nine in the evening and arrives in another city centre at eight in the morning uses time the traveller was going to spend asleep anyway.

None of this means the economics are easy. A sleeping car costs several times what an ordinary carriage costs and carries perhaps a third as many people. It earns money on one journey a day rather than on several, and it spends much of the daylight hours parked. Crossing a border adds further difficulty: signalling systems, electrical supply and safety rules still differ between countries, so a train that runs from Vienna to Brussels must be approved, and often re-crewed, in each country it passes through. Several announced routes have been delayed for years by exactly these approvals rather than by any shortage of demand.

Governments have begun to intervene, though unevenly. France has reopened routes it closed a decade earlier, with public money behind them, after a national debate in which the flight-versus-train comparison featured heavily. Sweden has subsidised a service to Germany. The European Union has funded studies and pilot routes, and at least one operator has ordered new sleeping cars — the first ordered in Europe for a generation — with delivery promised before the end of the decade.

Whether the revival lasts is a genuinely open question. The routes that have reopened are mostly the ones where the geography is favourable: a journey of nine to twelve hours, long enough that a night on board saves a hotel, short enough that the train arrives at a civilised hour. Journeys much shorter than that are better served by a fast daytime service; much longer ones struggle to compete with a flight at any price. Within that band, however, the night train has a straightforward advantage that no amount of aviation efficiency removes. It is the only form of long-distance transport that gives the traveller back the time it takes.

Passage 2 · Making concrete without the carbon

AConcrete is the second most used substance on earth after water, and its manufacture is responsible for something between seven and eight per cent of global carbon dioxide emissions. That figure surprises people who assume the problem lies with transport or electricity generation, and it is worth understanding where it comes from, because the answer determines which solutions can work.

BConcrete itself is mostly sand, gravel and water. The binder that holds it together is cement, and cement is where the emissions are. Making it requires limestone, which is crushed and heated in a kiln to around 1,450 degrees Celsius. Roughly 40 per cent of the resulting carbon dioxide comes from the fuel burned to reach that temperature. The other 60 per cent comes from the limestone itself: calcium carbonate decomposes into calcium oxide and carbon dioxide, and the gas is released whatever the kiln is heated with. This is the awkward part of the problem. A cement works powered entirely by renewable electricity would still emit the majority of what it emits today.

CThree broad approaches are being pursued, and they are not equally mature. The first is substitution: replacing some of the cement in a mix with an industrial by-product that behaves similarly. Fly ash from coal-fired power stations and slag from steel furnaces have been used this way for decades, and a mix containing thirty per cent slag performs perfectly well in most applications. The difficulty is arithmetic. The world produces far less of these by-products than it would need, and as coal power closes down, the supply of fly ash is falling rather than rising. Substitution can trim the total; it cannot solve it.

DThe second approach is capture. If the carbon dioxide is going to be released, it can in principle be collected at the chimney and stored underground or used to make something else. Several pilot plants now do this, and one Norwegian facility captures a substantial share of its own emissions. The engineering works. The obstacle is cost: capture adds considerably to the price of a tonne of cement, and cement is a commodity sold on price in a market where a few dollars a tonne decides a contract. Without either a carbon price or a purchasing rule that requires low-carbon material, a plant that installs capture equipment simply loses the business to one that does not.

EThe third approach is to change the chemistry. Several companies are developing binders that set without limestone at all, using magnesium compounds, alkali-activated materials, or calcium silicates that absorb carbon dioxide as they cure rather than releasing it. Laboratory results have been encouraging for twenty years. The difficulty here is not chemistry but time. A building is expected to stand for a century, and the only way to know whether a new binder lasts a century is to wait. Building codes reflect that caution: they are written around Portland cement, they specify it by name, and an engineer who substitutes an unfamiliar material takes on a liability that no commercial advantage justifies.

FThis is why the most effective interventions so far have had little to do with materials science. Several cities have rewritten their procurement rules to specify a maximum carbon intensity for the concrete in public buildings, which creates a guaranteed market for the low-carbon mixes that already exist and are already approved. Others have found that the cheapest saving is to use less: a structural engineer given the freedom to optimise a design can commonly remove a fifth of the concrete from a building without changing what it does, simply by placing material where the loads actually are rather than following a standard section. Neither measure requires anything to be invented.

GThe consensus among researchers is therefore unusually practical. The chemistry that would eliminate the problem may well arrive, but it will arrive slowly and it will be adopted more slowly still, because the industry's conservatism is a safety feature rather than a failure of imagination. Cement is not a consumer product that can be improved one model at a time; it is the material a bridge is made of, and the cost of being wrong about it is measured in lives rather than in returned goods. In the meantime the reductions available are the unglamorous ones: specify better, design leaner, and buy the material that already exists. The industry has spent a decade waiting for a breakthrough while declining to take the savings that were on the table, and the arithmetic of the next twenty years does not allow another decade of that. A twenty per cent reduction achieved everywhere by ordinary means would do more, sooner, than a ninety per cent reduction available only in the four plants that can afford it.

Passage 3 · Why experts disagree about expertise

In 1993 a paper appeared that has shaped the public understanding of talent ever since. Its authors studied violinists at a music academy in Berlin, sorted them by the judgements of their teachers, and reconstructed how much each had practised over their lifetime. The best students had accumulated markedly more hours of solitary, effortful practice than the good ones, and the good ones markedly more than the students training to be teachers rather than performers. The authors argued that what distinguished the groups was not an innate gift but the quantity and quality of what they called deliberate practice: work aimed squarely at what you cannot yet do, undertaken with full attention and with feedback, rather than the comfortable repetition of what you already do well.

The idea travelled a long way from the paper. In popular retellings it became the claim that ten thousand hours of practice will make anyone an expert at anything, a formulation the original authors never made and later objected to. The figure was an average for one group of musicians at one age, not a threshold; the paper reported wide variation around it; and the domain, classical violin, is one in which the criteria for good performance have been stable for two centuries and the training is unusually well codified.

Serious criticism began to accumulate. In 2014 a group of researchers re-analysed data from eighty-eight studies across music, sport, education, professions and games. They found that accumulated practice explained a substantial share of the difference in performance in games such as chess and a smaller though still meaningful share in music and sport, but almost none of the difference in professional work. The headline number — practice accounting for around twelve per cent of the variance across all domains pooled — was widely reported as a refutation. That reading goes further than the analysis supports. A variable that explains twelve per cent of the variance in an outcome is not a trivial variable; and pooling domains as different as chess and medicine produces an average that describes none of them.

The more careful objection concerns what the studies can and cannot show. Almost all of them are correlational: they observe that the better performers practised more, which is equally consistent with practice producing skill and with early skill producing the enjoyment, encouragement and opportunity that make long hours of practice possible. A child who finds the violin rewarding in the first year practises more in the second, and the two effects compound. Distinguishing them requires either an experiment nobody can ethically run or a long study that follows children before the difference emerges, and the handful of such studies that exist are small.

A second objection concerns the domain, and it is the one practitioners find most useful. Deliberate practice as originally described requires a settled idea of what good performance is, a teacher who can see the gap between the learner and it, and a task that can be repeated with feedback. Violin satisfies all three. Chess satisfies them. Surgery largely does. Management does not: the criteria shift, the feedback arrives years later and confounded by circumstance, and the same decision can be right in one firm and wrong in another. Where the conditions do not hold, the finding is not that practice matters less than we thought but that the word 'practice' does not describe anything specific enough to measure.

Both camps in the dispute now agree on more than the public argument suggests. Nobody serious claims that practice is irrelevant, and nobody serious claims any longer that ten thousand hours of anything guarantees anything. The live disagreement is about what else matters and by how much: working memory capacity, the age at which training began, the quality of the first teacher, physical characteristics in sport, and simple access to expensive instruction. Estimating the contribution of each is hard precisely because they are entangled with practice itself — a well-taught child practises better as well as more.

It is worth noticing what has happened to the disagreement itself. The original paper made a modest claim about one skill in one conservatoire; the public version made an immodest claim about everything; the critics then tested the immodest claim and found it wanting, and the result was widely reported as though the modest one had failed too. Something similar happens whenever a result travels from a journal to a bookshop, and it is a reason to read the finding rather than the headline — not a reason to distrust the research.

For a learner the practical conclusion survives the argument intact, which is perhaps why the original idea has proved so durable. Whatever proportion of expert performance practice explains, it is the proportion the learner controls. Innate differences cannot be acted on; the structure of an afternoon's work can. The specific advice that follows from the 1993 paper — work at the edge of what you can do, stay there long enough to be uncomfortable, get feedback that is specific, and do it again tomorrow — was never contingent on the ten thousand hours, and none of the criticism has touched it.