UC San Diego

The ability to produce accurately timed rhythmic actions is crucial for many of the most satisfying activities in life, like song, dance, and sport. Thankfully, most typical humans appear to perform wonderfully in colloquial and laboratory tasks that require them to synchronize movement with sounds and actions produced by social partners, environmental cues, or even themselves. As infants, children, and adults, humans also regularly learn new ways to interact rhythmically with the world, whether by playing clapping games as children or learning new partner dances or exciting new genres of music to perform or just nod along with. However, despite the apparent diversity in rhythmic actions and activities in daily life, many aspects of rhythm cognition have traditionally been viewed as highly stable. While we may learn to play a new rhythm on a guitar, the underlying building blocks of this rhythm are thought to be limited and shaped by mechanisms such generic properties of universally present neural oscillators or by an extremely pared-down memory for temporal intervals. These mechanisms are consistent with phenomena such as perceptual processing advantages for certain simple rhythms and the tendency for sensorimotor processes to warp the timing of complex rhythms to approximate the structure of simple rhythms. However, some support for the role of these mechanisms is called into question by recent evidence for widespread diversity not just in rhythmic activities broadly described but in fundamental aspects of the organization and distortion of time in rhythmic action and perception.In this dissertation, I consider the possibility that fundamental aspects of the sensorimotor maps which guide rhythmic action are plastic, in adulthood and across timescales. Through a series of audio-motor finger-tapping studies, I demonstrate that adults with limited experience in complex rhythms can rapidly improve their performance across the course of tens of minutes (Study 1). These rapid improvements are not solely due to nonspecific gains in attentional focus or sensitivity to timing feedback but are the result of learning processes which fine-tune and adapt the sensorimotor maps which transform rhythmic targets to executed rhythmic actions. This is evident from priming effects, where practice with different rhythms causes subsequent rhythms to be reproduced with different magnitudes and even directions of rhythm-warping (Study 2). It is also supported by the finding that gains in rhythm reproduction produced on one hand transfer across stretch breaks and to the other hand (Study 3). Longitudinal data recorded in near-daily intervals over the course of over a month also reveal that adaptations can build up across the timescale of weeks and lead to stable improvements in complex rhythm-performance. As a hint to the complexities of learning in the complex rhythm domain, recalibration that occurs on the scale of weeks was also found to be partially independent of that which occurs during a laboratory session, suggesting that multiple adaptive processes may operate in parallel. Combined, these findings paint a picture of human rhythmic action which is more flexible than previously envisioned.