Understanding nonhuman communication entails many challenges; largely because they exceed our aural capacity for sense making, but also due to the tendency to rely on our intuition about how sound and meaning align. Though we can’t understand what’s being communicated, we rely on our sensibilities about signification to parse the spectral properties of vocalization in other species. This problem is confronted in recent research on bat banter, a project that notably expands the range of social expression in non-primates.
Yosef Prat et al. devised a massive study of bats’ social sounds. They housed groups of Egyptian fruit bats (Rousettus aegyptiacus)—a very social and vocal species—in acoustically isolated chambers, monitoring them continuously with video cameras and microphones over 75 days. Of the tens of thousands of vocalizations recorded, they determined behavioral context, emitter’s identity, and the addressee for some 15,000 of these. The whole dataset covers the complete repertoire used by these bats over a period of 2.5 months. Most of what they recorded falls into the “agonistic” behavioral category in bioacoustics studies—squabbles over food, positioning in sleep clusters, mating attempts, or just getting in each other’s faces from being too proximate. That most of the vocalizations fall into this range is hardly surprising for a densely packed social species. The authors’ promote “studying the mundane, pairwise, directed, vocal interactions of animals,” which sounds quite similar to verbal social interactions of humans. But the distinctive feature of this study is that they settled on these categories by relying on an algorithm, one that was able to discern who was being addressed by the emitter. This results in two significant developments.
Typically, in bioacoustics studies on nonhumans, calls are separated into a-priori categories based on human-discernible acoustic studies. But the scope of bat sounds, and the vast quantities of data being generated, allowed the researchers to develop categories based on machine-learning analytics. From this shift in orientation, the subject of address then becomes intelligible. This is important because discerning forms of addressing individuals, aside from humans, has only been observed in dolphins and monkeys. In one regard, this makes them seem more like us: “our finding might be akin to a human speaker who uses varying intonation towards different listeners (e.g., male vs. female addressees) while using the same words.” But from another angle, the range of social subjects for modes of communication open up considerably. Instead of depicting creatures just squawking in the dark, bat sociality appears far more nuanced and complicated, with ongoing dialogues and disputes. The question now is whether the reliance on human-derived acoustic categories has been limiting our understanding of communicative patterns in other social species. Presumably so, and the authors’ critique of “seeking homology to semantics” will influence other researchers.
The broader relevance lies in the fact that bats are mammals. As the authors note, this provides as basis “to elucidate some of the huge potential of information capacity in a mammalian vocal communication system.” Given the confined settings of many bat communities, they could even become that rare taxa that serves as a model organism yet can be studied in the field, as well. If so, the subjectivity of these creatures models a more ample nonhuman social subject. The researchers speculate about what more they might hear: “The bat’s brain could thus be using some other representation that encapsulates much more information regarding different social aspects. The bat may be able to classify the context of an interaction with higher confidence, based on some acoustic feature which it evolved to use and is yet to be determined.” There’s so much more to listen for!
Yosef Prat, Lindsay Azoulay, Roi Dor, Yossi Yovel, “Crowd Vocal Learning Induces Vocal Dialects in Bats: Playback of Conspecifics Shapes Fundamental Frequency Usage by Pups,” PLoS Biology 15, no. 10 (October 2017): e2002556, https://doi.org/10.1371/journal.pbio.2002556.
Vocal learning, the substrate of human language acquisition, has rarely been described in other mammals. Often, group-specific vocal dialects in wild populations provide the main evidence for vocal learning. While social learning is often the most plausible explanation for these intergroup differences, it is usually impossible to exclude other driving factors, such as genetic or ecological backgrounds. Here, we show the formation of dialects through social vocal learning in fruit bats under controlled conditions. We raised 3 groups of pups in conditions mimicking their natural roosts. Namely, pups could hear their mothers' vocalizations but were also exposed to a manipulation playback. The vocalizations in the 3 playbacks mainly differed in their fundamental frequency. From the age of approximately 6 months and onwards, the pups demonstrated distinct dialects, where each group was biased towards its playback. We demonstrate the emergence of dialects through social learning in a mammalian model in a tightly controlled environment. Unlike in the extensively studied case of songbirds where specific tutors are imitated, we demonstrate that bats do not only learn their vocalizations directly from their mothers, but that they are actually influenced by the sounds of the entire crowd. This process, which we term "crowd vocal learning," might be relevant to many other social animals such as cetaceans and pinnipeds.