Medical News
Life
20 June 2019
Grey seals have been trained to mimic human sounds
By New Scientist staff and Press AssociationThree grey seals have been trained to copy speech, as well as notes from music including the Star Wars theme and the nursery rhyme Twinkle Twinkle Little Star.
Vincent Janik and colleagues at the University of St Andrews in the UK worked with the seals from birth, training them to copy new sounds by changing their formants. “Formants are emphasised frequency bands in our speech sounds,” says Janik. “They are parts of our speech sounds that we modify to encode information. For example, different vowels only differ in their formants.”
The seals were first trained to copy sequences of their own sounds, and then create melodies in their pitch. Human vowel sounds were later presented to the animals, which they then copied.
“It takes hundreds of trials to teach the seal what we want it to do, but once they get the idea they can copy a new sound pretty well at the first attempt,” says Janik.
Advertisement
One seal, named Zola, was particularly good at copying melodies played to her, including up to 10 notes of Twinkle Twinkle Little Star.
“Since seals use the same neural and anatomical structures as humans to produce these sounds, they provide a good model system in which to study how speech sounds are learned,” says Janik.
This isn’t the first time that seals have copied human vocalisations. A seal called Hoover was documented copying human speech – including phrases like “how are you?” – at the New England aquarium in Boston, Massachusetts, in the 1980s.
But this is the first study to really demonstrate how flexible seal vocalisations are, says Amanda Stansbury, who led the team. “Previous studies just provided anecdotal evidence for this.”
Journal reference: Current Biology, DOI: 10.1016/j.cub.2019.05.071
More on these topics:
animals
animal intelligence