Dispatch 02 — Barry, Wales
Your hearing fails from the top. Two industries are already there.
Part III names the model. A shop doorway in South Wales supplies the prototype. A receptor that wears out on a predictable schedule can be read as a specification, and in 2005 somebody read it that way.
FIGURES DRAWN FROM FOUR SEPARATE RECORDS · DEVICE SPECIFICATIONS AS STATED BY THE MANUFACTURER · AUDIOMETRIC PREVALENCE MEASURED AGAINST ADULTS WITH NORMAL STANDARD AUDIOGRAMS, NOT A GENERAL POPULATION · THE GLOBAL FIGURE IS POOLED PREVALENCE APPLIED TO POPULATION, NOT A COUNT
The Mosquito was built by Howard Stapleton in 2005 and first put to work outside a shop in Barry, South Wales. It emits a narrow tone near 17.4 kHz at up to 108 decibels. It does not detect loitering. It does not detect anything. It plays a frequency into a public space and waits for the people who can still hear it to leave.
The cochlea encodes frequency by position. The highest tones are read at the base, by the hair cells nearest the oval window, and those are the cells that go first — to age, to noise, to ototoxic drugs, in roughly that order of inevitability. Thresholds above 8 kHz start moving at around twenty-one in adults whose ordinary hearing tests clean, and early signs have been recorded in school-aged children from about nine. This is not a disease. It is the ordinary top-down wear of a receptor, and it is regular enough to build a product against.
Here is the part that makes the band useful to an industry and invisible to a clinic. A standard audiogram runs from 0.25 to 8 kHz and stops. That ceiling is not a finding about hearing. The Western Electric 1-A of 1914 tested to 16 kHz; the 2-A that replaced it in the 1920s stopped at 8, for reasons of calibration and instrument design, and the ceiling has held for a century. Everything the Mosquito uses sits above it. So does the earliest damage from headphones. A person can lose the top octave and pass the test.
By 2006 the transaction had been run in reverse. Students recorded the tone, loaded it as a ringtone, and took calls in classrooms where the teacher heard nothing. The same physics, sold twice, to both sides of the same room. That is what Part III means by a business model. The sense is the product surface, and the gap between who can hear and who cannot is the thing being monetised — in either direction.
Stage 01Age-graded decay
Hearing does not fade evenly. It fades from the top. Extended high-frequency thresholds begin to move at around twenty-one, and the order of loss is consistent enough across people to function as an age filter. Sense 03 does not merely get captured. It sorts its own users.
Stage 02The specification
A narrow tone near 17.4 kHz, up to 108 decibels, aimed at a pavement. Reported deployment runs to several thousand units; the published sales counts disagree with one another and are treated below as a range rather than a figure. The device sorts people by the condition of one receptor, which correlates with age, and by nothing else.
Stage 03The collateral
A frequency cannot read intent. Germany's Federal Institute for Occupational Safety and Health examined the device and declined to certify it safe, noting that at achievable levels dizziness, headache and nausea are to be expected, and that disruption of the equilibrium sense is a known effect of exposure in this range. Infants are the worst-placed listeners in the beam. They cannot leave, and the adult pushing the pram cannot hear what is happening. Sense 15 is collateral to a device aimed at Sense 03.
Stage 04The erosion
The headphone economy named in Part I works the same band from the other end. Pooled prevalence puts unsafe listening at roughly a quarter of young people through personal devices and near half through loud venues. Routine earphone use is associated with poorer extended high-frequency thresholds in adults who pass a standard test. The capture of Sense 03 shortens the very band Stage 01 depends on.
Stage 05The blind spot
The endpoint is not a clinic visit. It is the absence of one. Nineteen per cent of adults aged nineteen to thirty-eight carry measurable loss above 8 kHz behind a normal audiogram, and that loss tracks with needing a cleaner signal-to-noise ratio to follow speech in a room. The complaint is real, the test comes back clean, and the band holding the answer was ruled out of scope in the 1920s.
It is indiscriminate — babies, children, toddlers, autistic children, disabled children will hear this.
— Sir Albert Aynsley-Green, Children's Commissioner for England, 2008
That is the load-bearing sentence. A device aimed at a receptor cannot be aimed at a behaviour. Everything standing in the beam with an intact cochlea gets the same treatment, whatever it was doing there.
In June 2010 the Parliamentary Assembly of the Council of Europe adopted Recommendation 1930 unanimously. It found the dispersal of children and young people by these devices a disproportionate interference with rights under Article 8 of the European Convention on Human Rights, and asked all forty-seven member states to legislate against them. At the time of the recommendation, none had. Two years earlier the UK government had answered the Buzz Off campaign by confirming that the alarms were not banned and that it had no plans to ban them. Sheffield prohibited them on council property in January 2011.
What the record does not establish
- “Only under 25” is marketing, not audiologyThe tone sorts a distribution, not a birthday. Thresholds above 8 kHz vary widely between people of the same age. At 108 decibels plenty of adults over twenty-five perceive the tone, and plenty of people under twenty-five do not react to it. The device is a filter with a soft edge, sold as a switch.
- Hearing the tone is not a hearing testOnline “hearing age” pages are not audiometry. Phone and laptop transducers roll off near the top of the band, lossy codecs discard content there, and room noise masks what survives. A null result can be the equipment.
- Earphones: association, not causeThe extended high-frequency studies relate routine earphone use to thresholds. They do not establish a dose, a safe level, or that any individual's loss came from their headphones.
- The prevalence figure has a specific denominatorNineteen per cent is drawn from adults aged nineteen to thirty-eight whose standard audiogram reads normal. It is not a general-population rate, and it is not a rate of deafness.
- The global estimate is modelledThe range of 0.67 to 1.35 billion comes from pooled prevalence across thirty-three studies of roughly nineteen thousand people aged twelve to thirty-four, applied to population. Nobody counted a billion ears.
- Failure to certify is not proof of injuryThe German assessment declined to declare the device safe and described effects expected at achievable levels. That is a regulatory finding about an unassessed risk, not a measured injury rate at a stated exposure.
- The sales counts disagreePublished figures run from roughly three and a half thousand in use across England, to nine thousand sold in the UK, to six thousand worldwide, depending on source and year. The deployment was large. The exact number is not established by the public record.
The off-ramps that exist before the audiogram
The band cannot be restored once it goes. Everything below is about the part that has not gone yet, and about getting the loss measured while it is still information.
DeviceA weekly dose, not a volume knob
The WHO and ITU standard H.870 defines the allowance as exposure over time: 80 dBA across forty hours a week for adults, 75 dBA for children. Phones already track headphone level and already offer a hard output cap. Set the cap once. A limit enforced by the device survives the evening. A resolution does not.
VenueFiltered plugs, not foam
Foam cuts the top of the spectrum hardest, which is both the part already at risk and the part carrying intelligibility, so the room turns muddy and the volume goes back up. Flat-attenuation plugs drop the whole spectrum by a fixed amount and leave the mix intact.
ClinicAsk for the frequencies above 8 kHz
Extended high-frequency audiometry exists, runs from 9 to 16 kHz, and is established practice in ototoxicity monitoring. It is not part of a routine appointment unless it is requested. Trouble following speech in a noisy room alongside a clean audiogram is the exact presentation this band was left out of.
Public spaceThe device is a decision somebody made
Local authorities can prohibit them on their own property, and some have — Sheffield in January 2011. Council of Europe Recommendation 1930 remains the standing position of the Assembly. A tone in a doorway is a procurement line, not weather.
A field guide is not a clinical service. Hearing that drops suddenly, or drops in one ear, is an appointment within days rather than a protocol.
Sources
Srikanta K. Mishra, Udit Saxena, Hansapani Rodrigo. “Extended High-frequency Hearing Impairment Despite a Normal Audiogram: Relation to Early Aging, Speech-in-noise Perception, Cochlear Function, and Routine Earphone Use.”
Ear and Hearing 43(3), 822–835, 2022.
doi.org/10.1097/AUD.0000000000001140
Lauren K. Dillard, Malachi Ochieng Arunda, Lucero Lopez-Perez, Ricardo X. Martinez, Lucía Jiménez, Shelly Chadha. “Prevalence and global estimates of unsafe listening practices in adolescents and young adults: a systematic review and meta-analysis.”
BMJ Global Health 7(11):e010501, November 2022.
doi.org/10.1136/bmjgh-2022-010501
Parliamentary Assembly of the Council of Europe, Recommendation 1930 (2010), “Prohibiting the marketing and use of the ‘Mosquito’ youth dispersal device,” adopted 25 June 2010, rapporteur Piotr Wach.
assembly.coe.int
ITU-T Recommendation H.870 (V2, 03/2022), “Guidelines for safe listening devices/systems,” developed with the World Health Organization.
itu.int
Device specifications, deployment figures, the 2006 ringtone reversal, the February 2008 Buzz Off campaign and the January 2011 Sheffield prohibition are as reported in the contemporaneous public record. The safety assessment is that of the Bundesanstalt für Arbeitsschutz und Arbeitsmedizin, which did not certify the device as safe. The 8 kHz audiometric ceiling and its origin in the Western Electric 1-A and 2-A audiometers are as set out in Mishra, “Extended High Frequency Hearing Loss,”
AudiologyOnline, 13 June 2022.
◆
Dispatch 01 · Ontario
Dispatch 02 · Barry, Wales