The article that started it all: Oster 1973 on what binaural beats are for

Nearly every binaural beats product on the market traces its authority to one citation: Gerald Oster, Scientific American, October 1973. We read it. It is a well-written essay about how the brain combines sound from two ears, and it does not say what it is usually cited as saying.

Oster’s subject is mechanism. Play two slightly different tones through a loudspeaker and you hear ordinary beats, which are physical wave interference. Play them separately to each ear through headphones and you hear something else, which exists only because the two signals meet inside the head. He sets out the evidence that this is a different phenomenon: binaural beats are only audible when the carrier tone is low, becoming indistinct above about 440 Hz and vanishing by roughly 1,000 Hz (no listener he tested reported them above 900 Hz, though he notes others report up to nearly 1,500 Hz); they stay equally strong however mismatched the two tones’ loudness, and remain audible when one tone is a hundredth the loudness of barely audible sound; and unlike almost everything else in hearing, they are easier to hear against background noise rather than masked by it. He locates the processing at the superior olivary nucleus, the first point in the auditory path that receives input from both ears, and connects the whole thing to how we locate low-pitched sounds by interaural phase.

The original data in the article is a set of averaged evoked-potential recordings Oster made with Adam Atkin and Neil Wotherspoon at the Mount Sinai School of Medicine, using 300 Hz and 303 Hz tones. Monaural and binaural beats produced potentials that differed in amplitude, waveform and timing, which he reads as evidence that the two are processed at different sites. He reports no numbers, no statistics and no participant count, and he is candid about how fragile the recordings were: listeners had to concentrate on a faint, boring stimulus in total darkness, and often experienced auditory hallucinations that spoiled the trace.

What Oster proposes binaural beats are for is diagnosis. Examining neurological patients with his students, he found a few who could not hear the beats, some of whom also could not locate a finger snap, and some of whom had Parkinson’s disease; he describes one violinist who could not hear beats on admission and recovered the ability up to about 650 Hz over a week of treatment. Citing data from J. A. Tobias, he notes that women’s beat perception varies across the menstrual cycle while men’s does not, and suggests the beat spectrum might index hormonal state. His closing argument is that binaural beats deserve attention as a measurement tool, because they reveal processes the listener is not conscious of.

Limitations. Judged as evidence rather than as an essay, this is thin: a magazine article, no sample sizes, no statistics, no controls, no blinding. We grade it weak, and the grade is not a criticism of Oster, who was describing a perceptual phenomenon rather than testing a treatment. The point worth carrying forward is what is absent. There is nothing here about brainwaves being pulled into step with the beat frequency, nothing about relaxation, focus, sleep, meditation or altered states, and no therapeutic claim of any kind. Those ideas were attached to this citation later, by other people. When a product page says its approach is “based on Oster’s landmark 1973 research”, that research is about audiology and neurological examination.