Music, it is often said, is a central achievement of Homo sapiens. Rhythms have the power to bring people together, promoting social bonding and unlocking creativity. One of the central aspects of most modern music-making is the sense of cooperativity between musicians: one person will listen to and respond to the musical ideas of another, creating a dynamic and exciting song. However, despite its importance, the science behind this musical coordination is still unclear. How exactly do musicians listen to one another, integrate this information, and use it to alter the music they are playing, all in real time? Alternatively could music be mostly an autonomous, stereotyped affair which only seems to bear the hallmarks of a coordinated activity? Thankfully, science is starting to provide some answers here, thanks to a landmark paper by Eric Fortune and others in Science this month. The paper finds similar musical cooperativity in the duets of songbirds, expanding our understanding of how universal music is; and it also uncovers some of the exciting neural mechanisms through which this coordination is achieved.
Fortune’s research team investigated the cooperative aspects of music using a species of wren in which males and females sing coordinated duets. His research team wanted to determine whether wren duets were the product of a stereotyped, fixed pattern of vocalizations, or alternatively whether each wren could alter its singing in response to the syllables being sung by the other wren. This is fascinating because it might indicate whether the wrens genuinely coordinate their duets, or whether the duets simply appear coordinated but are actually the result of largely fixed individual behaviors. First, the researchers recorded hundreds of hours of wren song in the plain-tailed wren’s natural habitat of Ecuador. They captured many examples of duet singing, but also some examples of males and females singing their own portions of the duet alone, without a companion. In the solo recordings, the wrens left pauses between its syllables where the other wren should sing. However, these pauses were much longer than the intervals between syllables during a true duet; and moreover the pauses varied in length substantially compared with duets, where the pauses were of similar length. These key differences between solo and duet singing indicated that each wren might not have a stereotyped part it replays, but instead might change its song based on what its partner was singing during a real duet. In other words, coordinated music-making may be a real possibility in wrens.
The behavioral observations led the authors to examine the activity of individual neurons in the wren brains. They hoped to determine exactly how coordination happens on a neural level. The idea was this: a part of the wren brain had been identified which seemed to control song production (the HVC nucleus). Individual neurons in this nucleus could either be more responsive to the wren’s own portion of the duet, or to the partner’s portion of the duet, or to the entire duet performed by both individuals. This would provide some indication of whether the HVC nucleus simply produced the wren’s portion of the duet irrespective of what the partner was singing (if neurons responded most strongly to recordings of only the wren’s portion) or whether instead the HVC nucleus coordinated singing with the partner wren to produce a duet (if neurons responded most strongly to recordings of the full duet).
To this end, the researchers made audio recordings of duets performed by a number of wren pairs. They then captured the pairs and electrically recorded the activity of individual neurons as they played back either the entire duet, or just the male portions or female portions of the duet. They found that the majority of neurons in the HVC nucleus responded most strongly to recordings of the entire duet, even compared to the summed response of the male and female portions. This is strong evidence that individual wrens have intricate neural mechanisms responsible for listening to their partner’s song and using this information to change their own song, with the goal of producing a graceful duet. Also, when the researchers presented birds with male-only or female-only recordings in which the intervals between syllables had been tweaked to be outside the normal range, most neurons showed a decreased response. This is tremendously important since it indicates the wrens may have neural machinery which is very sensitive to the specific temporal progression of a duet. The speed of music, it seems, matters even to wrens.
Despite the fact that most neurons responded most strongly to the full duet, the researchers still found small populations of neurons which responded predominantly to either the female-only or male-only portions of the duet. This makes intuitive sense because the wrens need some way to not only coordinate singing, but also recognize their own portions of the song and their partner’s portions of the song. It’s still unclear how the different subtypes of neurons integrate all of this information to produce a song which can be dynamically altered as the partner sings. One way of solving this would be to map out the connections between areas of the HVC nucleus in order to see how the entire neural network responds to singing of a duet. This would allow researchers to determine how information is integrated to produce the very complicated behavior observed. The authors also hypothesize that counting neurons could keep track of the song’s beat, helping to coordinate behavior.
Another thought-provoking hypothesis mentioned by the researchers was that perhaps the females ‘led’ the duet. After all, when singing alone, males exhibited more variability in their singing than females and produced segments of lower volume than females. Furthermore, during duet singing, males sometimes missed their syllables, and in response the females lengthened the interval between their syllables and continued singing, possibly to lead the males to sing the right notes. Finally, in both males and females, the female portion of the duet elicited stronger activation of neurons than the male portion. Therefore the wrens not only coordinate singing, but the female may in fact lead the song!
Fortune’s team of researchers discovered some vitally important facts about musical cooperation which may be shown to be applicable in humans as well. It is likely that plain-tailed wrens coordinate their duet singing, responding to each other’s vocalizations on a syllable-by-syllable level rather than simply repeating rote behavior independently of their partner’s vocalizations. The HVC nucleus of the brain contained a majority of neurons which preferentially activated to the duet recording rather than either male-only or female-only recordings, providing some evidence of a neural mechanism for this. And finally, female wrens may in fact lead the duet song. While this may not be directly applicable to cooperative music-making in humans, it provides some terrifically important, elementary insights into how coordination between individuals is vital to music, and how neural circuits might orchestrate this coordination.
Article by Michael Lynn, 4th year BSc. Biology at the University of Ottawa.




