Band 8+ Academic Reading Practice Test 3

There is no such thing as a silent place on the inhabited surface of the earth, and the interesting question is not whether human noise reaches somewhere but for how long it is absent. Acoustic ecologists have adopted a deliberately modest measure: the length of the longest interval, during daylight, in which no human-generated sound is audible. In much of the developed world that interval is now measured in seconds.

The measure was chosen because the obvious alternative fails. Average sound level, expressed in decibels, describes the loudness of an environment and says nothing about its structure, and two places with identical averages can differ entirely in whether an animal can hear an approaching predator. A continuous low hum and an intermittent loud noise can produce the same number while doing quite different things to the organisms living there.

The effects on animals were documented first because they were easiest to isolate. Birds in noisy environments sing at higher frequencies, start earlier in the morning and, in several species, sing more loudly, all of which carry costs that have been measured. Species that rely on quiet listening rather than on calling — owls in particular — decline near roads at distances where the noise is barely noticeable to a walking human.

A study design that removed the usual confound settled a long-standing objection. Roads bring noise, but they also bring light, chemicals, collisions and the removal of habitat, so any effect near a road might have nothing to do with sound. Researchers built what they called a phantom road, playing recorded traffic through speakers in undisturbed forest with no road present, and found a substantial fall in bird abundance that reversed when the speakers were switched off.

Human effects are harder to establish and are not negligible. Nocturnal noise disrupts sleep architecture at levels well below those that wake a sleeper, and the resulting associations with cardiovascular outcomes survive adjustment for the usual confounders in several large cohorts, though not in all of them. The mechanism is plausible and the effect sizes are modest, which is a combination that supports precaution rather than alarm.

What makes the problem awkward is that quiet is not distributed like other environmental goods. It cannot be delivered to one household without being delivered to a neighbourhood, it is destroyed by the least considerate user rather than by the average one, and the people with least of it are, in every study that has looked, the people with least money. Regulation has concentrated on peak levels because they are enforceable, which leaves the measure that matters — how long the quiet lasts — untouched.