Birds and humans can learn songs and use them to engage in social communication. The learning process uses forebrain motor areas that give control over vocal equipment. The brain's cortex sends signals to vocal motor neurons in the brain stem. The bird or human also learns to depend on feedback sounds that help to develop the vocalizations that were learned and to maintain them. This is the process described in a published research report in Plos One.
The researchers were Gustavo Arriaga from the Department of Neurobiology at Howard Hughes Medical Institute, Duke University Medical Center in Durham, North Carolina. Erich D. Jarvis worked at Tulane University School of Medicine in New Orleans, Louisiana. Eric P. Zhou worked at both institutions.
It was believed that mice and non singing avian species lacked the physical mechanisms that singing birds and humans have, particularly "forebrain system for vocal modification". It was assumed that mice used ultrasonic sounds that they did not have to learn. This is called an "innate" characteristic where the mind controls a mechanism and the animal does not learn through experience.
Researchers found that mice have " a motor cortex region active during singing, that projects directly to brainstem vocal motor neurons and is necessary for keeping song more stereotyped and on pitch." The researchers also found that "...male mice depend on auditory feedback to maintain some ultrasonic song features, and that sub-strains with differences in their songs can match each other's pitch when cross-housed under competitive social conditions." In other words, the researchers concluded that mice do have some of the neuroanatomical features that birds and humans have and that they do learn to sing in similar ways as humans and birds that are equipped to sing.
It was known that male mice used an ultrasonic vocalizations when they were mating. These vocalizations were like those used by songbirds, meaning that melodic structures, a sequential vocal structure, syllables that were not in a random order, a short rhythmic or melodic passage, and differences in each mouse's collection of vocalizations. Again, songs with this level of complexity were assumed to be innate.
Part of the experiment involved deafening some mice and leaving others with their hearing ability. The animals were allowed to sing, then they were killed so their brains could be examined. The deaf mice still did the innate singing, but had differences in activity in their primary auditory cortexes. The researchers tested for and found features that are less developed than in humans and songbirds, but included, "... forebrain activation, direct cortical to vocal motor neuron connectivity, forebrain control, auditory feedback, and vocal imitation".
Some important discoveries were described by Brian Owens in Newsblog. He notes that "The researchers discovered that mice do have a brain circuit that starts in the primary motor cortex, projects directly to the part of the brainstem responsible for controlling the larynx, and importantly, is active when male mice sing. The difference, when compared with birds and humans, is that the circuit is weaker, more sparse. It’s there, it’s just not as strong."
He also highlights the idea that different strains of mice have their own sounds or "accents". Another difference is between singing and non-singing mice is that, when a singing animal goes deaf, its songs deteriorate. But if a mouse is congenitally deaf, it's songs are of low quality to begin with. The deterioration in song quality does not happen in mice or other species who use innate vocalizations. In other words, auditory feedback is needed for an animal to maintain learned songs."
The presence of these features was found to support the hypothesis that mice actually learn to sing in a way that is like humans and songbirds and that, while their physiological mechanisms are not as well developed, they are not just letting their brains do the work on an innate basis.