Spontaneous, predictive, tempo-flexible synchronization to external rhythms is a unique phenomenon that has been observed in humans and only a handful of parrot species, all of whom display a capacity for vocal learning. According to the vocal learning hypothesis (Patel, 2006), the ability to synchronize movements to an external auditory beat in a predictive and tempo-flexible way is linked to vocal learning, i.e. the capacity to learn complex vocalization patterns based on an auditory model. Horses (Equus ferus caballus) may present an interesting challenge to this hypothesis. They are not vocal learners, yet there are many anecdotal reports of horses displaying beat synchronization-like behavior. This claim however has not been empirically investigated; the purpose of the present study is to do so.

To determine whether horses spontaneously synchronize movement to an external beat in a predictive, tempo-flexible manner.

Eight horses stabled at the Mount Holyoke College Equestrian Center were tested using the circular trotting to music paradigm described by Bregman, et al. (2013). Horses were held on a lead line and lunged at a trot in a circle while musical excerpts were played over arena loudspeakers. Excerpts from two songs (“Happy” and “Shake It Off”) were played at their canonical tempo (160 bpm) and four shifted tempi (144, 152, 168, 176 bpm); these tempi were selected due to prior report of 152 bpm as a typical trotting rate (Gallo, 2007). Each horse trotted to twenty 45-second excerpts, grouped into two ten-excerpt blocks. Before each block, horses trotted in silence for 60 seconds. Horses changed direction between blocks. The lunger wore earplugs and listened to a dense, arrhythmic audio mixture over noise-cancelling headphones to avoid delivering rhythmic cues to the horses. A second researcher delivered visual cues to the lunger when excerpts started and stopped, but otherwise remained out of sight. Horse gait was recorded using high-speed video and accelerometers mounted to forelimb boots. Forelimb hooffalls were identified, and inter-hooffall intervals (IHIs) were calculated to assess trotting synchronization to the musical beat.

A range of preferred inter-hooffall intervals was observed across horses (preferred IHI: 343 – 411 ms; 146 – 175 bpm). However, IHIs were not modulated by musical excerpt tempo (p > .7), demonstrating that horses do not spontaneously engage in predictive, tempo-flexible synchronization to a musical beat.

Despite anecdotal claims to the contrary, horses do not spontaneously synchronize their trotting to musical beats in a predictive, tempo-flexible manner. This finding is consistent with the claims of the vocal learning hypothesis that such synchronization is linked to neural structures underlying vocal learning.