The enjoyment of music is ubiquitous across human societies and cultures. Among the (bio)cognitive underpinnings to process and enjoy music, rhythm plays a key role. In humans, musical beat processing intimately links perception and action when we entrain rhythmic movements to musical beats. In social settings, this leads to rhythmic actions within groups of people, such as dancing or marching in unison, but what selective pressures led to rhythmic behaviour to begin with, and why did the social use of rhythm evolve? The search for the origins of social rhythm is complicated because unlike other biological traits, rhythmic processing does not fossilize and humans only constitute one datapoint to build testable hypotheses on rhythm evolution. However, rhythmic processing is not unique to humans, with examples found across the animal kingdom. In this project, we will integrate approaches from field biology, comparative neuroscience, artificial intelligence, and speech sciences to test competing hypotheses on the evolutionary roots of rhythmic abilities. We will study a wide range of marine mammal species, known for their vocal flexibility but subject to differing social pressures, as a test-bench for evolutionary hypotheses on the origins of social rhythm in our own species.