Of all the world's animals, only humans, some kinds of birds and perhaps some porpoises and whales learn the sounds they use to communicate with each other through a process of listening, imitation and practice. For the rest, including nonhuman primates, these sounds develop normally in the absence of external models.
Now Rockefeller University scientists have found that zebra finches, songbirds native to Australia, use infant-like strategies to learn their song. Some finches focus on perfecting individual song components, referred to as "syllables," while others practice longer patterns called motifs. Which strategy they choose, or what combination of strategies, seems to depend on what their siblings are doing. In time, all are able to sing the same adult song.
The results, reported in the December 13 online issue of Proceedings of the National Academy of Sciences, are the first to show a social influence on how birds learn their song by analyzing song-learning with birds kept in family groups rather than in isolation chambers.
The Rockefeller team also shows for the first time that individual birds, of the same species, can follow different strategies to get to the same end point of singing the adult song. Until now, scientists thought that the vocal learning process in birds was mainly a matter of filling in details in a pre-existing developmental program. If so, then this program is, in zebra finches, a very flexible one.
"This research points to a remarkable parallel in vocal learning in infants and some songbirds" says senior author Fernando Nottebohm, Ph.D., Dorothea L. Leonhardt Professor and head of the Laboratory of Animal Behavior at Rockefeller.
"In both cases vocal learning seems to be approached as a challenge in problem solving," says Nottebohm, whose studies in canaries in the 1980s provided the first evidence of spontaneous neuronal replacement in the adult vertebrate brain.
A problem-solving approach may apply to other kinds of sensory motor learning beyond vocal learning, he added, suggesting that zebra finches may offer further insights into human learning.
"I find it amazing that something that infants, with brains weighing approximately 1,000 grams, do over a period of years can be accomplished, perhaps in a similar way, by young songbirds over a period of weeks, with brains weighing just 1 gram," says Nottebohm.
"Of course," he adds, "the diversity of sounds mastered by the young birds is much smaller, but all the same there is a remarkable parallel between what they do and the way in which humans acquire the sounds of language."
To learn to sing, a young male zebra finch (only the males sing) first has to hear and memorize the song of a nearby adult male. This begins to happen about 20 days after the bird hatches. By 35 days of age, he begins to imitate. At first he produces what scientists call a subsong -- considered analogous to babbling in human infants.
The young male sits, eyes closed and feathers ruffled, and burbles a stream of variable sounds that are soft and rambling in comparison to the louder, very structured, adult song. When he makes these soft sounds, he is not communicating to other birds. Gradually the syllables of the final song become more recognizable. By 80 to 90 days of age, the bird is both sexually mature and sings the song he will use in courtship. After that he sings that song exactly the same way every time and will do so for the remainder of his life. Zebra finches can live for up to 8 years.
The Rockefeller University studies are the first to investigate whether young birds take different approaches to mastering the sounds they imitate. "No one had thought a lot about strategies," says Nottebohm. "How, starting from an unstructured beginning, do you approach the final stereotyped song?"
To investigate this question, Wan-chun Liu, Ph.D., a postdoctoral fellow and first author of the paper, and Nottebohm studied 37 young male zebra finches from 15 clutches that were kept in cages that they shared with their parents and siblings. In nature zebra finches are highly social birds that breed in colonies of up to several hundred pairs. A young bird learns its song by imitating his father or other adult males, often copying different parts of the song from different adults.
Liu observed the birds and recorded them on tape, at first every other or every third day. Between days 35 and 50, as the young birds began imitating, they were observed for six or seven hours a day. "It's very difficult when more than two birds are singing together to tell which bird is producing which song," says Liu. To do so, he spoke into the microphone to identify which bird was singing.
The recorded songs were then analyzed with a computer program that produces a sound spectrogram -- a visual representation of the sound that plots frequency over time. This allowed the researchers, with help from Timothy Gardner, Ph.D., another postdoctoral fellow, to quickly see similarities and differences among the songs. Adult male zebra finches sing only one song, a mixture of scratchy and nasal sounds clustered into several distinct "syllables," with each syllable preceded and followed by a brief silent interval. The entire song lasts about one second. "It's very brief and unassuming and not particularly musical," says Nottebohm, "but practical to quantify."
By the time the birds were 43 days old, two clear strategies of imitation were apparent. About half the birds tended to repeat one song syllable many times; from these repetitions all of the syllables in the adult song eventually emerged. This was the repetition strategy. The others attempted to sing the entire song motif, with all its different syllables, like their adult model, and did so in a way that was noisy and imprecise. This was the motif strategy. Of the latter birds some included the silent intervals between the different syllables and others jumbled all the sounds together without interruptions. In one group of three siblings, each bird followed a different strategy although all were imitating the same adult song, that of their father.
To better quantify their results, the Rockefeller scientists used computer software that analyzes pitch, frequency modulation, tonality on a scale from pure tone to white noise, and spectral continuity, which allows integration of the information over time. This analysis allowed them to compare and quantify the similarities between sounds, how the sounds changed over time and the extent to which they eventually matched the adult song being imitated.
Liu also observed 12 additional birds that were removed from their parents and siblings when 20 to 25 days old and housed individually in soundproof chambers with one adult male zebra finch to imitate. These birds used both the syllable-repetition and the motif strategies to learn their song, but combined the two more often than the birds raised with their families. Regardless of the strategy followed, all the birds mastered the imitation at about the same age.
The researchers propose that in a family setting, a young zebra finch chooses a strategy different from that of his siblings, perhaps to better track his own vocal development as he learns the song.
Earlier studies of vocal learning in zebra finches have focused on single birds housed alone imitating a recorded song. "Nobody followed what happens if you put several birds together," says Liu. "The diversity of learning styles became apparent when we studied the birds under conditions closer to nature. They try to avoid each other's style of song learning."
"This study suggests that social interactions greatly affect the way a song is learned," he adds.
Human infants also follow different routes toward mastering the sounds of language, for reasons that remain unknown. Some infants focus at first on repeating individual words and others go through a stage of short, jumbled phrases, mostly unintelligible, with the cadence and inflection of adult speech. Eventually the individual words become clear.
"In both infants and zebra finches vocal learning does not unfold in a pre-set manner, but rather emerges as an exercise in problem solving that leaves much room for external influences and individual learning styles," Nottebohm says. "We're not teaching our zebra finches how to learn their song -- how to get there is totally up to the birds."
This research was supported in part by the National Institutes of Health, the Mary Flagler Cary Charitable Foundation, the Herbert and Nell Singer Foundation and the Phipps Family Foundation. Liu also was supported by a Li Memorial Scholar Fund fellowship.
Cite This Page: