This fascinating research suggests some languages are more dangerous than others.
- Droplets travel huge distances from sneezes and coughs, but also from everyday speech.
- Fast, vowel-heavy languages likely transmit more moisture than slower languages with less airflow.
Scientists believe the way we speak could affect rates of transmission of COVID-19 (coronavirus)—and some languages are better than others, because they require less of the airflow that pushes moisture droplets out of our mouths. But does this really make a difference?
One of the key ways even homemade cloth masks are thought to help slow the spread of coronavirus is by simply keeping your moisture to yourself. Even humid regular air carries an amount of moisture in the form of microscopic droplets, but while air molecules pass easily through a cloth barrier, water droplets are more likely to get snagged.
Scientists have focused a lot on transmission of droplets by coughing, which is a common symptom of what is, at the most basic level, a respiratory illness. And sneezing always makes the news because its trajectory and power are surprising, like bottle rockets or lightning strikes for wild fast particles that come out of your nose and mouth.
But language is a bit of a sleeper agent for airborne moisture.
In English, only a few sounds can be made without any airflow, and these are consonants that mean little without supporting vowels. Sibilant or strident sounds like ss, sh, th, and f even make you blow more air through a smaller mouth shape—anecdotally, the times you’re most likely to accidentally spray moisture onto your conversation partner.
Vice shares a popular video where someone speaks in Japanese and then English. “When the woman speaks English, the phrase causes the cloth to flap in the wind emitted from her mouth; when she speaks Japanese, it stays nearly entirely still,” Vice reports. Even phonemes that look similar on paper may be “aspirated” more or less in different languages.
The stats do appear to back up the language transmission theory. As of press time, Japan has had 17,292 confirmed coronavirus cases and 920 deaths; the U.S., meanwhile, has had 2.05 million confirmed cases and 115,000 deaths.
What’s happening? The sibilants and stridents are like the sneezes of English vocalizing, in that they’re obvious choices for spraying the most moisture. Researchers have dug deeper to study how the entire spectrum of vocal sounds includes surprising differences in droplet spray.
“Case reports of transmission at choir practices, restaurants, and call centers suggest that it is possible for SARS-CoV-2 to hitch a ride on aerosol particles and stay suspended in the air,” Vice reports. Scientists have been assembling and studying different sentences to see which produce the most droplets.
In the summer of 2003, when the number of probable SARS cases in Japan remained zero and other countries, like the U.S., had more confirmed cases, Japanese researcher Sakae Inouye shared his hypothesis in The Lancet: Different languages, because of their vowels and other airy sounds, could have different rates of infection following normal conversations.
“With such large numbers of visitors from Japan and USA, why have no Japanese visitors contracted the virus?” Inouye wondered, and speculated that different levels of aspiration between English, Chinese, and Japanese could be partly responsible.
In addition to the basic ideas of more and less aspirational sounds, some languages are just more information-dense per spoken syllable. Mandarin Chinese, for example, is packed with sibilant and strident sounds like English, where air is pushed with more force through the mucus membranes. But each syllable can be spoken with four different tones or the “neutral” tone, changing the meaning without adding more sounds or breath.
Indeed, a 2011 study ranking major languages by “density” of syllable information put Mandarin at the top for density. And languages with lower density, like Spanish, were found to be spoken so quickly that the syllable density matters less in the big picture. The researchers suggested this is why some languages sound so fast compared to others.
Information-dense Mandarin has one of the slowest syllable speeds. And when the issue is the amount of airflow, a lower-density language with higher speed is pushing out more air that could contain droplets.