Tongue twisters, such as the well-known tale of Peter Piper and his pickled peppers or the image of a woman selling sea shells by the seashore, can sound quite different when attempted while sober compared to after consuming alcohol.
Currently, scientists theorize that alterations in pitch and frequency, specifically, could serve as indicators of one’s level of inebriation.
According to Dr. Brian Suffoletto, the primary researcher of a Stanford University study, this method has various possibilities for future use. One potential application is as an ignition lock on vehicles, preventing someone from starting their car unless they can successfully complete a “voice challenge.” This could also be implemented in certain high-risk workplaces such as school bus drivers or heavy equipment operators to ensure the safety of the public.
Another potential use for this technology could be in restaurants or bars, allowing bartenders to be alerted when it may be necessary to stop serving a customer.
In a study published in the Journal of Studies on Alcohol and Drugs, Suffoletto and his team describe how they observed 18 adults over the age of 21 attempting to recite a challenging phrase.
Following the consumption of a high dose of alcohol, each individual was instructed to recite a different tongue-twister every hour for up to seven hours. The participants’ speech was then recorded.
Suffoletto stated that the tongue-twister was used to test vocal stress and reveal changes that may not be noticeable when speaking regular prose.
The scientists also recorded participants’ breath alcohol levels prior to consuming the beverage and at 30-minute intervals afterwards until seven hours after drinking.
The rhymes were recorded in one-second segments and the vocal characteristics of pitch and frequency were examined. A subset of the data, along with corresponding breath alcohol levels, was used to train an AI system. The remaining voice data was then used to test the system.
According to the findings, the voice analysis was able to predict alcohol intoxication with a 98% success rate. Intoxication was determined as having a breath alcohol concentration of 0.08% or higher, which is the legal limit for driving in the US.
However, there are limitations to the study, as it only focused on white participants and did not consider other aspects of speech that may be affected by alcohol, such as volume.
According to Suffoletto, effectively using harm prevention messages also involves strategic timing. For example, reminding someone of their drinking limits when they start drinking can have a strong impact. However, once they are heavily intoxicated, the effectiveness of these interventions decreases.
The study was conducted by Petra Meier, a public health professor at the University of Glasgow. She was not directly involved in the research, but she emphasized that the study was small and had strict controls in place.
She stated that there is potential for exciting advancements that could potentially be highly beneficial. However, it would be necessary to first test this method on a larger and more diverse group. She also emphasized the importance of testing the approach in real-life scenarios.
Meier stated that it is important to ensure that the method works effectively for both light drinkers and frequent, heavy drinkers. This is because heavy drinkers may not exhibit noticeable changes in their voice until they have reached significantly higher levels of intoxication, which could lead them to falsely believe they are capable of driving safely.