IELTS Academic Reading Practice Test 122

In the years after the American Civil War, a saloon owner in Dayton, Ohio, faced a problem familiar to every retailer of the period. His staff served drinks and took payment, but he had no way of knowing whether the money that reached him matched the number of drinks poured. In 1879 James Ritty, with his brother John and a mechanic named John H. Eckert, produced a machine intended to settle the question. According to the company's later accounts, the idea came to Ritty on a steamboat, where he noticed a device that recorded the revolutions of the propeller. If turns could be counted, he reasoned, so could sales.

The brothers were granted a patent for what they called the Ritty Incorruptible Cashier in 1879. The machine was built of brass and iron on a wooden base. Across its front ran a row of keys, one for each coin denomination and some for larger sums, and beside them a lever. When a sale was made, the assistant pressed the appropriate keys and pulled the lever. A bell sounded, a small window displayed the amount, and the drawer sprang open. The bell was not a courtesy to the customer: it was a signal to the owner, who could hear it from another room, that a transaction had been recorded.

Two features made the machine difficult to cheat. The amount of the sale was displayed in a window that the customer could see before the drawer opened. More importantly, every sale was added to a running total shown on a dial, so that the day's takings could be checked against the money actually in the drawer. A coin could not be removed and the drawer closed again without the display, the total and the ringing bell testifying to what had happened. The register did not merely hold money; it produced a record.

Ritty had little capital and no experience of manufacturing. He sold his share of the business, and in 1882 the rights passed to a group who formed the National Manufacturing Company. Two years later the enterprise was reorganised as the National Cash Register Company, under John H. Patterson, a coal dealer who had bought a controlling interest. Patterson was not an engineer, but he understood something that Ritty had not: the value of the machine lay less in the hardware than in the reassurance it gave an employer.

Patterson built one of the most formidable sales organisations in American business. Agents worked on commission and were set quotas, and those who failed were dismissed. A salesman was expected to approach a shopkeeper with a demonstration rather than an argument, and to leave with a signed contract and a deposit. The company ran its own training school and published a magazine for its staff. Visitors to Dayton were shown not only the factory but the town's schools, parks and hospitals, an early exercise in corporate public relations.

Competition was handled severely. NCR bought rival patents, undercut smaller firms and, according to the courts, pressed shopkeepers to return or destroy machines made by competitors. In 1913 Patterson and other executives were convicted under the Sherman Antitrust Act of conspiring to restrain trade and to drive competitors out of business, although the verdict was later overturned on appeal. The case did not destroy NCR, which continued to dominate the market for decades, but it made the company's methods a byword for aggressive selling.

Meanwhile the machine itself grew more capable. By the 1890s registers could print each transaction on a paper roll, providing a permanent record that an assistant could not alter, and could separate the takings of different departments within a store. Special versions appeared for hotels, department stores and banks. The bell, the drawer and the printed slip became so familiar that the American phrase for making a sale, 'ringing it up', is generally thought to come from the sound the register made.

The company also employed remarkable engineers. Charles Kettering, who joined NCR in 1904 and later became celebrated for the electric starter that made hand-cranking a car unnecessary, worked on an electrically driven register. A motor replaced the crank, so that the drawer opened and the totals were calculated at the touch of a key, and the machinery became quieter and faster. Electric registers spread through shops during the first half of the twentieth century, but the mechanical design changed little.

The revolution came from electronics. From the early 1970s gears and levers were replaced first by electronic circuits and then by microprocessors, which could store prices and calculate tax. In June 1974 a packet of chewing gum in a supermarket in Troy, Ohio, became the first item sold with the help of a printed barcode. The register was no longer an isolated machine but part of a system linking the checkout to the stockroom and to the company's accounts.

The deeper legacy of the cash register is a change in the nature of shopkeeping. Before Ritty's machine, a shop's takings were a matter of memory and trust; afterwards they were a set of figures that an owner could inspect and compare. Patterson's product gave employers a means of supervising their staff, and it gave the shop something new: data.

For most of the history of satellite communications, the machines have been placed as far from the Earth as practical. A satellite in geostationary orbit circles at about 35,800 kilometres above the equator, completing one revolution in the time the Earth takes to turn once, so that it appears to hang motionless above a fixed point. Three such satellites can cover almost the whole planet. The arrangement is elegant, but it has drawbacks, and since about 2019 a different approach has dominated investment: the mega-constellation, in which hundreds or thousands of satellites fly in low Earth orbit, between roughly 300 and 1,200 kilometres up, and hand the work from one to the next as they pass overhead.

The main reason for coming down is delay. Radio waves travel at the speed of light, but light takes time to cover distance, and a signal that goes up to geostationary orbit and back covers more than 70,000 kilometres. That journey alone accounts for about a quarter of a second, before any processing is added, and the delay is noticeable in a video call or an online game. At 550 kilometres the round trip is a few thousandths of a second, short enough for engineers to call the connection low-latency. The difference matters for financial trading, for remote medicine and for any application in which two machines must respond to each other in real time.

Low orbits bring a second advantage: frequency reuse. A geostationary satellite illuminates a large area with a limited number of beams, whereas a dense constellation can point narrow beams at small cells and use the same frequencies many times over. The cost is complexity. Each satellite at 550 kilometres is above the horizon for only a few minutes, and the terminal must switch its link from a satellite that is setting to one that is rising, a process called handover, without the user noticing. Software in the terminal and in the network's control centres manages this continuously, and in principle the loss of a single satellite is almost invisible to users, because its neighbours take over the traffic within seconds.

The route a message takes is not the one most people imagine. The subscriber's terminal, a flat panel containing hundreds of small radiating elements, forms a beam that it steers electronically towards the satellite in view, with no moving parts. The satellite receives the signal and sends it down to a gateway station, a large dish connected to the internet by fibre-optic cable. When the satellite is over an ocean, however, there is no gateway within reach. For this reason the newer constellations carry laser terminals that allow satellites to pass data directly to one another, so that a message can hop across the Pacific and come down where a cable reaches land.

Radio spectrum is a shared and finite resource, and the growth of constellations has made its management contentious. Operators must obtain licences from national authorities, and the radio regulations of the International Telecommunication Union, a United Nations agency, allocate frequency bands to different services and register each satellite network in order to avoid interference with its neighbours. Most broadband constellations use the Ku and Ka bands, which allow high data rates but are weakened by heavy rain. Optical links between satellites avoid both problems: there is no rain in space, and a laser beam needs no share of the radio spectrum.

What the systems are used for has changed as capacity has grown. The first customers were households in rural areas where no cable had ever been laid. Ships, aircraft and military units are now important users, since a moving terminal can track a satellite across the sky, and after hurricanes and earthquakes portable dishes have restored communications for rescue teams within hours. The economics, however, are uneven. Launching is the largest single cost in any constellation, and at 550 kilometres the tenuous atmosphere still produces enough drag to pull a dead satellite down within about five years. Demand is highest in cities, where cables already offer faster and cheaper service, so the constellations find their market in places that are awkward to reach.

Astronomers were the first group to raise a serious objection. A satellite crossing a telescope's field of view leaves a bright streak on the image, and even a satellite too faint to be seen by the eye adds a background glow that must be removed before the data can be used. Surveys that photograph the whole sky repeatedly are particularly vulnerable, because the same patch may be crossed many times in a night. Operators have responded by darkening the surfaces of their spacecraft and by steering them away from the most sensitive instruments. Astronomers acknowledge the cooperation but point out that the sky is now crowded in a way it was not a decade ago.

The more serious long-term risk is collision. Satellites and debris move at nearly eight kilometres a second, and a collision at that speed would shatter a spacecraft into thousands of fragments, each capable of destroying another. Operators therefore track every object larger than a few centimetres and move their satellites out of the way; one company reported tens of thousands of avoidance manoeuvres in a single year. The nightmare scenario, described by the American scientist Donald Kessler in 1978, is a chain reaction in which debris generates further debris until an orbital shell becomes unusable. Space weather adds a further hazard: in February 2022 a geomagnetic storm increased the drag on 49 newly launched satellites and 38 of them were lost.

Few subjects have entered the primary curriculum as quickly as computer programming. In England the national curriculum for computing, introduced in 2014, made programming and algorithm design compulsory for children from the age of five, and several other countries have since followed. Inexpensive devices and free environments such as Scratch, developed at the Massachusetts Institute of Technology, allow a seven-year-old to build a working game in an afternoon. The enthusiasm is easy to understand. Yet the case for teaching coding to young children rests on several distinct arguments, and they are not equally strong.

The first argument is economic. Coding, we are told, is the literacy of the twenty-first century, and children who learn it will be equipped for the jobs of the future. This is the weakest of the arguments. Most of today's primary pupils will do work that no one can now describe, and the specific languages and tools they learn at seven will be obsolete long before they look for employment. Programming is also increasingly assisted by software that writes much of the routine code. If schools justify a subject by its supposed effect on employment they will always be training children for a world that has already moved on.

The second argument is about thinking. Seymour Papert, who helped to develop the Logo programming language in the late 1960s, believed that programming a computer gave children a powerful way to think about their own thinking, and the phrase computational thinking now appears in almost every curriculum document. The claim is that learning to break a problem into steps, to test a solution and to find the mistake in it develops a general capacity that will show up in mathematics and science. It is an attractive idea, and it may yet prove true, but reviews of the research have found the evidence for such transfer to be inconsistent and often weak. Measuring it is genuinely difficult, because children who learn to code differ in many other ways from those who do not.

The third argument is the strongest, and it is rarely the one that governments make. Digital systems now decide what a child sees, what a child is offered and, increasingly, how a child is assessed. To understand that these systems are built by people, that they contain assumptions and errors, and that they can be examined rather than merely obeyed, is a form of civic education. Programming is not the only way to teach it, but it is a particularly direct one: a child who has made a computer do something ridiculous has learnt that the machine is not magic.

Whether any of this happens depends on teachers. Primary teachers are generalists, and most were trained before computing existed as a school subject. Many are anxious about it, and anxious teachers fall back on recipes: pupils follow a set of instructions, produce an identical result and learn little. Teaching programming well requires a different approach, in which mistakes are the material of the lesson rather than a failure, and in which a class is given time to explore. That requires sustained training and, above all, time out of an overcrowded timetable.

There is also a question of fairness. Where computing is well taught, it tends to be in schools with confident staff and reliable equipment; where it is not, it becomes an optional club for the children whose parents can provide a laptop at home. A child who owns a computer and has an adult who can help has a considerable advantage over one who has neither. The subject risks reproducing the inequalities it is supposed to reduce. Countries that have introduced coding most successfully have treated it as part of a broad entitlement, with equipment and training provided centrally rather than left to individual schools to find.

None of this means that coding should be dropped. The time available in a primary school day is finite, and every new subject competes with reading, arithmetic and play, but programming can be integrated with those rather than added to them: a child who writes a program to draw a regular polygon is doing geometry, and a child who animates a story is doing both English and art. The worst outcome would be a separate, test-driven subject in which children learn to reproduce code they do not understand.

Programming also suits young children better than its reputation suggests. It is concrete, it produces something visible, and it rewards persistence. Short, frequent sessions with physical devices, paired work in which one child writes and the other checks, and unplugged activities that teach the underlying ideas with paper and cards have all been used successfully. What does not work is a long weekly lesson in a computer room that the class visits as a treat. Play is not the opposite of rigour; in the early years it is usually the route to it.

My conclusion is therefore a qualified one. Coding deserves its place in the primary curriculum, but as a branch of a broad education in computing rather than as vocational training, and its advocates should be honest about what is not yet known. Schools should promise less on its behalf: not a boost to every other subject, not a guaranteed career, but a way of helping children to see that the digital world is made by people, and can be understood, questioned and changed. That is a modest claim, and it is one the evidence can support.