Unlearned calls are produced by all birds whereas learned songs are only found in three avian taxa, most notably in songbirds. The neural basis for song learning and production is formed by interconnected song nuclei: the song control system. In addition to song, zebra finches produce large numbers of soft, unlearned calls, among which “stack” calls are uttered frequently. To determine unequivocally the calls produced by each member of a group, we mounted miniature wireless microphones on each zebra finch. We find that group living paired males and females communicate using bilateral stack calling. To investigate the role of the song control system in call-based male female communication, we recorded the electrical activity in a premotor nucleus of the song control system in freely behaving male birds. The unique combination of acoustic monitoring together with wireless brain recording of individual zebra finches in groups shows that the neuronal activity of the song system correlates with the production of unlearned stack calls. The results suggest that the song system evolved from a brain circuit controlling simple unlearned calls to a system capable of producing acoustically rich, learned vocalizations.

Summary To understand both proximate and ultimate factors shaping vocal communication, it is fundamental to obtain reliable information of participating individuals on different levels: First, it is necessary to separate and assign the individuals' vocalisations. Secondly, the precise timing of vocal events needs to be retained. Thirdly, vocal behaviour should be recorded from undisturbed animals in meaningful settings. A growing number of studies used animal‐attached microphones to tackle these issues, but the implications for the study species and the research question often receded into the background. Here, we aim to initiate a discussion about the limitations, possible applications and the broader potential of such methods. Using lightweight wireless microphone backpacks (0·75 g including customised leg‐loop harness) combined with multi‐channel recording equipment, we captured vocal behaviour of small songbirds. We evaluated the effect of the devices at various levels, including an assessment of how vocal and locomotor activities were affected by initial device attachment and battery exchange. We compared our approach to existing studies and identified suitable research examples. We acquired continuous vocalisation recordings of zebra finches, and unequivocally assigned them to interacting individuals, with system‐based synchrony, irrespective of background noise. We found effects of initial backpack attachment and of battery replacement on vocal and locomotor activity, but they were minimised through the extended recording duration (ca. 16 days) that outlasted habituation effects (ca. 3 days). This method provides the tools to integrate individual vocal communications into a group setting, while enabling animals to behave freely in undisturbed, structured and acoustically complex environments. By minimising the effects on the animals, the behaviour under study, and ultimately on the research question, this approach will revolutionise the ability to capture individual‐level vocalisations in a variety of communication contexts, opening up many new opportunities to address novel research questions.

Many organisms coordinate rhythmic motor actions with those of a partner to generate cooperative social behavior such as duet singing. The neural mechanisms that enable rhythmic interindividual coordination of motor actions are unknown. Here we investigate the neural basis of vocal duetting behavior by using an approach that enables simultaneous recordings of individual vocalizations and multiunit vocal premotor activity in songbird pairs ranging freely in their natural habitat. We find that in the duet-initiating bird, the onset of the partner’s contribution to the duet triggers a change in rhythm in the periodic neural discharges that are exclusively locked to the initiating bird’s own vocalizations. The resulting interindividually synchronized neural activity pattern elicits vocalizations that perfectly alternate between partners in the ongoing song. We suggest that rhythmic cooperative behavior requires exact interindividual coordination of premotor neural activity, which might be achieved by integration of sensory information originating from the interacting partner.