Think back on the last time you had a conversation with someone: how often were both people talking at the same time? Unless your conversation was a heated argument, chances are that the majority of the time you and the other person took turns speaking, overlapping only rarely.
Turn-taking is a universal phenomenon across all human languages and appears early in infancy, suggesting that it is a fundamental feature of human communication. Increasingly, turn-taking has also been found to take place across a variety of animal species. It is likely so widespread because it affords communicational benefits, enabling information to be exchanged between parties in an orderly and efficient fashion without interference. However, taking turns also requires some mechanism of coordination among participants.
Research on turn-taking in animals has mainly focused on how communicating pairs coordinate. However, many social species including humans live and communicate in larger groups. As anyone who has ever attended a dinner party or a work meeting knows, coordinating turn-taking can become much more challenging in larger groups.
Photo credit: Douglas Branch
In a recent study, we investigated group calling sessions in meerkats, a social mongoose species that lives in groups of 2 - 50 individuals. On cold mornings, meerkats typically spend some time warming up together in the sun outside their sleeping burrow. During these sunning sessions, they often produce short vocalizations known as “sunning calls”. These calls are given primarily when other meerkats are around, and rarely when meerkats are sunning alone.
By analyzing audio recordings of meerkat groups taken in the Kalahari Desert, South Africa, we found that meerkats tended to avoid interrupting one another during these group sunning sessions. We first measured how often the calls from a given meerkat overlapped with calls from others in its group. We then compared these measured overlap rates to simulated overlap rates generated by shifting that meerkat’s calls forward or back by a fraction of a second or a few seconds relative to the calls of its group mates. We found that the overlap rate was minimized when no time shifting was applied (i.e. in the natural calling data), and dramatically increased for even small simulated time shifts. In other words, meerkats strongly avoided overlapping one another’s calls.
To understand how meerkats managed to avoid overlap, we looked at how individuals timed their calls relative to the calls of their group mates. We found that meerkats tended to suppress their calling immediately after hearing a group member vocalize. However, after this initial period of call suppression (which lasted for about 200 milliseconds), they actually increased their call rates upon hearing the calls of others. This finding suggests that turn-taking in meerkats is driven by two processes that occur over different time scales: over the short-term the calls of others cause a meerkat to suppress its calls, but in the long-term they stimulate it to call more. In combination, these two processes allow meerkats to carry on prolonged vocal exchanges while also avoiding overlapping one another’s calls.
Overall, our results show that turn-taking can be maintained in vocal interactions involving more than two parties, and that it may be driven by relatively simple behavioral interactions. While the specific function of turn-taking in meerkat sunning interactions is unknown, evidence from other species suggests that it might go beyond information exchange. Coordination can often play a role in strengthening social bonds, and it is not unlikely that vocal exchanges during sunning could also serve this purpose for meerkat groups. In the future, we hope to further explore the function of turn-taking, as well as to test whether turn-taking extends beyond the context we studied to other aspects of social living.
Reference
Vocal turn-taking in meerkat group calling sessions. Demartsev, V.*, Strandburg-Peshkin, A.*, Ruffner, Michaela, & Manser, M. (2018) Current Biology 28(22):3661-3668. *equal contributors.