For decades, scientists believed that humans and a handful of musical birds held a near-exclusive claim to one of the most culturally defining abilities on Earth: the ability to move in time with a beat.
However, a surprising new study now suggests that our sense of rhythm may not be as uniquely human as once thought.
In a series of experiments, macaque monkeys were observed spontaneously tapping along to music, locking onto rhythms even when it offered them no reward.
Published in Science, the study reveals that monkeys can not only perceive a steady beat in real music, but can also anticipate future beats and adjust their tapping accordingly—all skills once thought to emerge primarily in vocal-learning species such as humans and songbirds.
This unexpected finding challenges a long-standing hypothesis about the origins of musicality and opens a brand-new window into how the building blocks of rhythm may have evolved.
“Synchronizing movements to music is a hallmark of human culture, but its evolutionary and neurobiological origins remain unknown,” the researchers write. “Here, we demonstrate that macaques can synchronize to a subjective beat in real music and even spontaneously do so over alternative strategies.”
A new challenge to a long-held theory
For years, the dominant explanation for why humans can keep a beat has been the “vocal-learning hypothesis.” The theory proposes that rhythmic synchronization evolved as a side effect of complex vocal learning. Under this view, species like humans, parrots, and songbirds can lock onto a beat because they possess advanced vocal-mimicry abilities, while most mammals—including our closest primate relatives—cannot.
However, this new study’s findings disrupt that clean evolutionary story.
Two adult macaques, previously trained to tap in time with metronomes, were introduced to something far more complex: real music. Unlike metronomes, songs lack clean, repeating onsets and pose a much greater cognitive challenge. Yet the macaques learned to extract and follow the beat—sometimes even preferring to tap in sync with the music despite easier options.
The findings suggest that rhythm perception and synchronization may exist on a continuum across species—one that doesn’t require advanced vocal mimicry as a prerequisite.
How do you teach a monkey to “feel” the beat?
In the study, researchers placed monkeys in front of a screen equipped with an infrared tap sensor and an audio playback system. Once they initiated a trial by touching a holding bar, a song would start, prompting them to start tapping.
Three musical excerpts with well-established human tapping consensus were used, each presented at a distinct tempo. To earn a reward, the monkeys needed only to maintain consistent intervals between taps—not to tap at any particular phase relative to the music.
This distinction is crucial: the monkeys were never trained to tap on the beat. Yet, that’s exactly what they did.
Even more impressively, when the researchers shifted the onset of the song’s audio—misaligning the cue that normally triggered tapping—the monkeys’ taps shifted accordingly.
In humans, this kind of “phase shift” response indicates genuine beat tracking rather than a rote reaction to visual signals. In the monkeys, the same behavior suggested they were following something in the music itself.
“The monkeys were never trained or rewarded to produce a particular phase,” researchers wrote. “Despite this, both monkeys produced a consistent tapping phase for all three musical exercises.”
To test whether monkeys were really hearing and responding to rhythm—rather than simply learning a behavior—the researchers introduced scrambled music.
They chopped the original songs into tiny 30-millisecond fragments and rearranged them, destroying the temporal structure while preserving the same acoustic frequencies. The result was a burst of noise with no beat.
When listening to scrambled music, the monkeys still tapped consistently to receive rewards—but their taps no longer aligned with the audio, suggesting that synchronization emerged only when the stimulus contained meaningful rhythmic structure.
“With the exception of one scrambled song for each animal, the tapping phase distributions now did not differ between the original 0 and π versions of these scrambled stimuli… indicating that although monkeys were perfectly capable of ignoring the auditory stimulus,” the researchers explained. “They chose to synchronize to some feature in the stimulus when the temporal structure was informative.”
This was one of the clearest signs that macaques weren’t just performing a trained behavior—they were actually listening.
You must be logged in to post a comment.