As a species, we can produce fluid speech that can elicit an emotional response. This is a godsend for the species in general, but it is a burden for individuals with problems with their larynx.
For those who are either born unable to talk or who lose their speech later in life, using a robotic device can only help so much.
Stephen Hawking and his voice synthesizer is one of the most notable cases of a human using a prosthetic voice box.
But though the physicist was able to speak any word in the dictionary, his machine could not place emphasis on certain words or give the impression that Hawking was asking a question rather than simply saying a statement.
A group of researchers has recognized the potential for a device that could give those without a functioning voice box a realistic-sounding voice.
The team — based at the University of California, San Francisco — has started the first steps toward such a discovery by determining which area of the brain is responsible for controlling pitch in human speech.
Pitch allows us to convey mood or emphasis when speaking. High and low pitches are created by the vibration of vocal cords. These, in turn, are controlled by tension in the folds that comes from flexing muscles, causing a faster vibration.
The study, which was published in the journal Cell, aimed to see which brain areas were activated when different pitches were created. This could only be done by studying people’s brains as they speak in real time.
Luckily, senior author Dr. Edward F. Chang is a neurosurgeon who works with a number of epilepsypatients. Some of his patients’ brains are fitted with minuscule electrodes that can help to determine the origin of a seizure.
Using the same detection method, which is known as electrocorticography, the scientists recruited volunteers from this group of epilepsy patients.
They were all asked to repeat the following sentence: “I never said she stole my money.” Each time, participants were instructed to place emphasis on a different word in order to change the meaning of the sentence (and the pitch of their voice).
By closely studying brain activity, the team found that neurons in one particular brain area — the dorsal laryngeal motor cortex — were activated when pitch was altered. When the pitch became higher, the area showed more activity.
As well as examining activity, researchers also electrically stimulated neurons in this area. This caused the larynx muscles to flex and even prompted a vocal response in some patients.
Participants also experienced listening back to their own voices, which caused a response in the dorsal area.
The results could help scientists to figure out how the human brain manages to mimic the voices of others.
Of course, this will go some way toward explaining how the brains of impersonators work, but it has greater importance; it could help to engineer a prosthetic voice box that gives patients a natural voice, allowing them to speak in a more realistic and less monotonous way. “You’d really like a prosthetic to carry the emotional content of the speaker. It’s not enough to just capture the words because so much of what we communicate is how we say something.” Lead author Benjamin Dichter
The researchers have not finished their work just yet. Along with studying how such a prosthesis could be developed, they are also looking into “reverse engineering” the brain’s pitch control.
This means attempting to predict which word is being emphasized just by studying the brain’s neural activity. One thing is clear: there is a lot more to learn about the human voice.