When a horse whinnies, it produces two sounds simultaneously. The low-frequency component — around 200 Hz — comes from vibrating vocal folds, the same mechanism that produces most mammalian vocalizations, including human speech. The high-frequency component — above 1000 Hz — is a whistle generated inside the larynx, produced by airflow through a constriction rather than tissue vibration. Two completely different physical mechanisms, running in parallel, encoding different emotional information in the same vocalization.
Reader note (2026-11-19): The Current Biology paper (36, 902-911, 2026, DOI 10.1016/j.cub.2026.01.004) is titled 'The high fundamental frequency in horse whinnies is generated by an aerodynamic whistle'; its byline is Lefèvre, Barluet de Beauchesne, Sabarros, Briefer Freymond, Ramseyer, Keller, Reby and Fitch. I read its abstract through Europe PMC (the full text was not reachable); the abstract gives a fundamental frequency above 1,000 Hz in addition to a lower one of about 200 Hz, as the essay says, and describes biphonation and its likely evolution 'to convey multiple independent messages concurrently', with no mention of arousal or valence. The assignment of the two bands to arousal and valence is the finding of an earlier paper, Briefer et al., Scientific Reports 5, 9989 (2015), whose abstract says that all whinnies contained two fundamental frequencies, 'F0' and 'G0', 'not harmonically related, suggesting biphonation', and that 'F0 and the energy spectrum encoded arousal, while G0 and whinny duration encoded valence'. The essay presents the dual mechanism and the information split together as what the February 2026 paper found; the 2015 paper reports the two frequencies and the arousal and valence split, and the 2026 title is about the mechanism that produces the high one.
The dual mechanism was discovered through acoustic analysis published in Current Biology in February 2026. Small rodents produce laryngeal whistles. Large mammals vibrate vocal folds. Horses do both, at the same time. Each frequency band carries different information: the low-frequency component correlates with arousal level, the high-frequency whistle correlates with emotional valence — whether the experience is positive or negative. A single whinny communicates both how excited the horse is and whether it's happy or distressed.
The structural insight is about bandwidth through multiplexing. A single vocalization has limited information capacity if it uses one physical mechanism — the pitch, duration, and modulation of a vibrating-fold sound can encode only so much. By running a second mechanism in parallel on a different frequency band, the horse doubles its communication bandwidth without doubling the time spent vocalizing. This is frequency-division multiplexing — the same principle that allows fiber optic cables to carry multiple signals on different wavelengths of light.
The evolutionary question is why horses evolved this and other large mammals did not. Horses are social prey animals that need to communicate across open grassland while remaining alert for predators. Long vocalizations attract attention. Short vocalizations carry less information. Multiplexing solves this tradeoff: compress two messages into one burst. The selection pressure is not for louder or longer calls, but for denser ones.
This suggests that communication capacity is not determined solely by cognitive complexity. Horses are not considered among the most cognitively sophisticated mammals, yet their vocal communication is structurally more complex than most primates'. The complexity is in the channel, not the mind. When the environment demands dense signaling under time pressure, evolution builds a wider pipe.