Beat keeping in a Sea Lion as Coupled Oscillation: Implications for comparative understanding of human rhythm

Interest in whether the ability to pick up on and synchronize with a beat in musical rhythm is uniquely human or may be more widespread throughout the animal kingdom is on the rise. In fact, non-human animals’ musical abilities were the topic of a dedicated symposium at the International Conference on Music Perception and Cognition earlier this year (“Music Perception & Cognition Across Species”, featuring talks on chimpanzees, several bird species, and Ronan the sea lion [admittedly not all of which were necessarily focused on rhythm and beat perception]). Most tests of non-human animals’ synchronization abilities have been conducted with metronomes or metronome-like stimuli, and some evidence for synchronization abilities has been reported for bonobos, chimpanzees, and budgerigars. Several papers have also documented synchronization to more complex rhythmic stimuli, like real music – however, it’s important to distinguish between flexible, anticipatory synchronization and “bouts of synchronization” that may reflect a transient phase alignment between two oscillators with similar tempi (which will eventually and transiently occur for any two uncoupled oscillators with similar but not perfectly matched periods). Moreover, if we’re to conclude that an animal or animal species is capable of human-like beat perception or synchronization, then direct experimental manipulations such as perturbations to the rhythm are necessary – that way, we can observe compensatory dynamics of the animal’s behavior in order to infer the properties of the underlying neural or behavioral oscillator that make synchronization possible. That’s exactly what a recent paper by Andrew Rouse, Peter Cook, Edward Large, and Colleen Reichmuth has done.

The paper focused on Ronan (a sea lion), who has previously been shown to be capable of synchronization (of head bobs) to the beat in real music, even when tempo shifted (excellent YouTube video available here: In the current paper, Ronan synchronized to metronome-like stimuli that contained phase or period perturbations of different magnitudes and in different directions (advance/delay, speeding/slowing). The authors examined the patterns of compensatory behavior exhibited by Ronan in response to these perturbations. Importantly, the design allowed the authors to fit Ronan’s data with nonlinear equations that describe the behavior of coupled oscillators (here the coupling is between Ronan’s head bobs and the metronome stimulus). The model estimates the extent to which Ronan needed to adjust the phase and period of her head bobs to resynchronize with the metronome following a perturbation.

Ronan’s behavior (and model fits to that behavior) revealed that Ronan flexibly synchronized with metronomes with different tempi (similar to her music synchronization performance). She was also able to adapt to both phase and period perturbations to the stimulus, getting back into sync within a handful of intervals. Similar to what is normally observed for humans, Ronan’s phase correction was stronger than her period correction (tempo adaptation). In contrast to humans though, who normally show relatively constant phase correction estimates (and even react to subliminal phase perturbations), Ronan’s degree of adaptation to phase perturbations scaled with the magnitude of the perturbation. Ronan also seemed to show weaker period correction than humans typically do. However, with respect to these two divergences from human literature, it is notable that Ronan’s performance was not directly compared to a human sample performing the same task with the same stimuli.

Overall, the experimental design (introducing perturbations in rhythmic stimuli) and approach to analyzing the data (involving fitting coupled-oscillator models) are, in my opinion, a model of where investigations into non-human beat-keeping abilities should be going. To my mind, the only missing piece is the direct comparison to a human sample, as it’s unclear whether discrepancies between Ronan’s performance and previous human data are fundamental or might stem from something uninteresting like the nature of the stimulus (which for Ronan is filled with a cycle of a frequency modulation, diverging from the types of stimuli that are often used in human studies). Regardless, the similarities between Ronan’s performance and typical human performance certainly outweigh the differences, and provide further evidence that beat perception and synchronization may not be specific to humans or to vocal learning species. Instead, the ability to synchronize in a flexible, anticipatory way might be more universal than once thought, potentially stemming from common neural circuitry that gives rise to oscillatory activity capable of synchronizing with environmental rhythms. Demonstrating this in different animal species may simply be a matter of identifying an appropriate behavior or task that allows a particular species to show off their skills, rather than expecting that a human-centric task like finger tapping will be the key to revealing cross-species similarities in beat perception abilities and music abilities more generally.

–Molly Henry, University of Western Ontario

Source article: