This paper proposes that music did not begin with the voice, with abstract cognition, or with the manufacture of instruments. It began with running. The biomechanics of persistence hunting forced the human body into sustained rhythmic entrainment — and the mathematical ratios the body discovered in that process are the same ratios that govern musical meter and consonant pitch across every culture in recorded history.
Contemporary theories of music's origin divide along two principal axes: the sexual selection hypothesis (Darwin, 1871), which frames music as a courtship display analogous to birdsong; and the social cohesion hypothesis (Dunbar, 2012), which positions music as a mechanism for reinforcing group bonds. Both theories account for music's social function. Neither accounts for its structure — specifically, for the universal preponderance of regular, hierarchical meter, and for the cross-cultural convergence on a small set of interval ratios (3:2, 4:3, 5:4) in melodic pitch systems.
This paper proposes a third, biomechanically grounded origin: music's structural properties were entrained into the human nervous system through the mechanics of long-distance running. Persistence hunting — the practice of pursuing prey at submaximal speed until its thermoregulatory collapse — required early Homo to sustain rhythmic locomotion for periods of 4–8 hours. This duration is sufficient for the body to undergo deep entrainment between its oscillatory systems: cardiac cycle, respiratory cycle, and locomotor cycle.
The key mechanism is Locomotor-Respiratory Coupling (LRC): the synchronization of breathing rhythm to footstep rhythm at precise integer ratios. As metabolic load increases with fatigue, the body cannot smoothly modulate these ratios — it must shift discontinuously between stable integer states. This paper argues that these discrete shifts are the body's first algorithmic time-signature changes, and that the ratios the body found stable (9:8, 5:4, 4:3, 3:2, 2:1) are precisely the ratios that define musical consonance and meter in every documented musical culture.
A secondary contribution addresses the relationship between pitch and rhythm. These are not categorically different phenomena — they are the same physical process (periodic oscillation) observed at different temporal scales. Below the 20 Hz perceptual boundary, individual cycles become distinct percussive events: a rhythm. The drum, this paper argues, was not invented; it was recognized — as the acoustic externalization of the human running pulse.
The Persistence Entrainment Hypothesis is an original theoretical contribution by Java Roque (Javier Roqueni), first proposed in 2026. The component facts cited throughout this paper — LRC biomechanics, persistence hunting anatomy, the 20 Hz perceptual boundary, and basal ganglia beat processing — are established science with attributed sources. The synthesis is new. To the author's knowledge, no prior published work has:
If you are a researcher aware of prior work that substantially overlaps with any of the above claims, the author welcomes correspondence at hola@javaroque.com. This theory is offered in the spirit of open intellectual inquiry.
In 2004, Bramble and Lieberman published a landmark study in Nature demonstrating that the human body is not a generalist primate body adapted secondarily to occasional locomotion. It is a specialized endurance running machine — a body whose anatomical configuration is explicable only by sustained selective pressure toward running performance over millions of years.
The morphological evidence is extensive. The nuchal ligament — absent in chimpanzees, present in all running quadrupeds — stabilizes the head during the ballistic phase of the running stride. The enlarged gluteus maximus, a muscle minimally active in walking but maximally active in running, anchors trunk stability during single-leg support. The shortened forefoot, widened shoulders (decoupled from the pelvis for rotational counterbalancing), elongated Achilles tendon, and expanded calcaneal tuberosity all appear together in the fossil record at approximately the Homo erectus transition (~1.8–2 million years before present) and are absent in Australopithecus.
Bramble and Lieberman's argument was this: these adaptations are inexplicable as walking adaptations. They only make sense as running adaptations. And the most probable selective context was not short-burst predation (for which speed, not endurance, would be selected) but persistence hunting: the tracking and pursuit of prey over hours, exploiting the unique human capacity for sweat-based thermoregulation against the prey's panting-based system, which fails first under sustained heat load.
The significance for music theory is chronological: if the anatomical adaptations for endurance running predate the Upper Paleolithic creative explosion (~40,000 BP) by approximately 1.8 million years, and if those adaptations entrained rhythm into the human nervous system through daily practice across thousands of generations, then the capacity for rhythmic entrainment was not a product of musical culture. Musical culture was a product of it.
Locomotor-Respiratory Coupling (LRC) describes the entrainment of respiratory cycle to locomotor cycle in running animals. In obligate quadrupeds (horses, dogs, cats), LRC is mechanically enforced: the visceral piston mechanism — the inertial displacement of abdominal organs during gallop — compresses the diaphragm on each footfall, making a strict 1:1 stride-to-breath ratio energetically unavoidable.
In humans, the constraint is softer. The bipedal gait decouples the respiratory system from the visceral piston. Nevertheless, experimental studies of human runners consistently show a strong preference for LRC at integer ratios. Bramble and Carrier (1983) identified preferential coupling at 4:1, 3:1, 2:1, and 3:2 (strides per breath cycle) during treadmill running across a range of intensities. Bernasconi and Kohl (1993) documented further stability at 5:2 and 4:3. The coupling is not obligatory — but it is preferred, and the preference is strong enough to be measurable across subjects running at different cadences and at different experience levels.
The critical insight is why the brain prefers integer ratios at all. The respiratory and locomotor oscillators are mechanically coupled through thoracic pressure dynamics. When their frequencies share a simple integer ratio, the energy cost of maintaining phase coherence between them is minimized: the systems reinforce rather than interfere. This is the same physical principle underlying musical consonance — a frequency ratio of 3:2 between two oscillators produces less destructive interference than a ratio of 3.17:2, because the former is periodic in time whereas the latter is not.
The brain, in optimizing metabolic efficiency during running, is solving the same problem it later solves when evaluating musical intervals: find the simplest integer ratio that satisfies the coupling constraint. The aesthetic preference and the physiological preference are the same computation.
As a persistence hunter's metabolic load increases over the course of a chase, the required breathing rate rises continuously. But the LRC ratio cannot rise continuously — it is quantized. The body cannot smoothly interpolate between a 20:20 ratio and a 14:13 ratio the way an electronic oscillator can sweep through frequencies. It must find the nearest stable integer ratio and lock onto it.
This produces a series of discrete transitions: a staircase of respiratory states rather than a ramp. Each transition is a gear shift in the body's metabolic machinery — and each gear corresponds to a distinct rhythmic relationship between breath and stride that the runner experiences as a perceptible change in their internal pulse. The body announces each shift. You feel it.
| Gear | Ratio (in:out) | Steps/Breath Cycle | Breaths/Min at 160 BPM | Musical Meter Equivalent | Musical Interval |
|---|---|---|---|---|---|
| Endurance | 20:20 | 40 | 4 | ∞ — No discrete meter | Unison |
| Gear 2 | 14:13 | 27 | ~6 | 27/4 — Searching, unstable | Minor Second (searching) |
| Gear 3 | 9:8 | 17 | ~9 | 17/8 — Proto-meter | Major Whole Tone (9:8) |
| Gear 4 | 7:6 | 13 | ~12 | 13/6 — Shifting | Septimal Minor Third |
| Gear 5 | 5:4 | 9 | ~18 | 9/4 — Compound meter | Major Third (5:4) |
| Common Time | 4:4 | 8 | 20 | 4/4 — Common Time | Octave Equivalence |
The table above encodes what this paper considers its central claim: the sequence of gears that a persistence hunter traverses as fatigue accumulates corresponds to a sequence of time signatures, moving from long, unstructured respiratory cycles through increasingly articulated rhythmic structures, arriving finally at 4/4 common time — the most prevalent meter in all documented musical traditions.
The gear shift is not merely a metabolic event. It is perceptible, it is patterned, and it is repeatable across individuals and across days. Early humans who ran together experienced these transitions simultaneously — synchronized not by convention but by physiology. The group breath-lock that occurs when runners entrain to each other's respiratory phase is the world's first ensemble performance, unrehearsed and unavoidable.
In standard acoustic pedagogy, rhythm and pitch are treated as categorically different phenomena. Rhythm belongs to the temporal domain; pitch belongs to the frequency domain. This distinction is phenomenologically real — we hear them differently — but it is physically unfounded. Both are periodic oscillation. The difference is entirely a matter of rate.
The human auditory system integrates periodic stimuli in time. Below approximately 20 Hz, the cochlear hair cells respond to individual pressure cycles as discrete events: we hear each cycle as a separate click, pulse, or impact. This is rhythm: the temporal perception of individual oscillation cycles. Above 20 Hz, the nervous system can no longer resolve individual cycles; it integrates them into a continuous percept characterized by a single quality — pitch. The 20 Hz boundary is therefore not a threshold between two different physical phenomena. It is a threshold between two different modes of neural processing of the same physical phenomenon.
The implication for percussion is direct. A drumstroke produces a broadband acoustic transient whose period is determined by the rate of repeated striking, not by the resonant frequency of the drumskin. A drum played at 160 BPM produces events at 2.67 Hz — far below 20 Hz, clearly in the rhythmic domain. Now consider: the running footstrike also produces a broadband acoustic transient at 2.67 events per second (160 BPM / 60). The footstrike and the drumstroke are acoustically identical in their temporal structure. The drum is not a novel invention. It is the oldest sound in human experience, rendered portable.
There is a further consequence: any pitched oscillator slowed below 20 Hz transitions from pitch to rhythm. A 440 Hz sine wave slowed by a factor of 22 becomes a 20 Hz flutter. Slow it by a factor of 200 and it becomes a 2.2 Hz pulse — a heartbeat. The pitch and the heartbeat are the same wave at different speeds. This is not a metaphor. It is a physical identity.
The most powerful evidence for the Persistence Entrainment Hypothesis is not any single piece of paleoanthropological or acoustic data. It is a convergence: the same small set of integer ratios appears, independently, in three completely separate domains of human knowledge.
The ratios are: 2:1, 3:2, 4:3, 5:4, 9:8. In music theory they are known as the octave, the perfect fifth, the perfect fourth, the major third, and the major whole tone — the five simplest ratios in Pythagorean and just intonation systems, forming the skeleton of every major and minor scale in Western music and their analogs in non-Western systems from Indian raga to Indonesian pelog.
These are also the ratios observed with greatest frequency in human LRC studies. And they are the ratios at which mathematical Lissajous figures — the geometric forms produced when two oscillators interact at a given ratio — close into recognizable, stable, non-chaotic geometric patterns.
This is not numerical coincidence. These ratios are small. The human nervous system, whether it is synchronizing two body oscillators during running, evaluating two musical pitches for consonance, or tracking the phase relationship between beat and melody, is performing the same fundamental computation: find the simplest integer ratio that minimizes phase error between two periodic signals. The brain optimizes for synchrony. Simple ratios produce synchrony. Music, at its structural core, is the formalization of the ratios the body discovered in motion.
Pythagoras discovered the perfect fifth by observing string lengths. The running body discovered it by optimizing breath. Both arrived at 3:2 because 3:2 is the simplest ratio that produces non-trivial coupling. Pythagoras had a lyre. The persistence hunter had a body. The persistence hunter came first by approximately 1,999,500 years.
The standard organological narrative positions the drum as an early human invention: a resonant surface struck to produce rhythmic sound for social, ritual, or communicative purposes. This framing treats rhythm as a mental concept that humans projected outward onto manufactured objects.
The Persistence Entrainment Hypothesis reverses this causality. The drum did not teach humans rhythm. Humans, having already internalized a deep rhythmic structure through millions of years of locomotion, encountered the drum as a recognition: here is an object that makes the same sound the body makes, at the same frequencies, with the same transient profile. The drum was not invented. It was named.
The acoustic evidence is precise. A human footstrike on packed earth produces a broadband impulse with energy concentrated between 50 and 300 Hz and a sharp attack transient in the first 5–20 ms. This is acoustically indistinguishable from a hand or mallet strike on a taut skin drum. The drumskin is tuned (through its tension and diameter) to resonate at frequencies in the same 50–300 Hz range. The drum does not merely imitate the footstrike — it amplifies and clarifies it. It makes the internal rhythm of running audible at a distance.
There is a secondary implication for body percussion. The earliest percussion instruments may have been the body itself: thigh-slapping, chest-striking, foot-stamping — all of which produce broadband transients in the same frequency range and are documented as pre-instrumental percussion practices in numerous ethnomusicological traditions. The body was the first drum. Ground-striking was the first stick.
If the Persistence Entrainment Hypothesis is correct — if musical meter is the cultural formalization of locomotor entrainment — there should be a shared neural substrate for locomotor timing and musical beat processing. There is.
The basal ganglia are a set of subcortical nuclei long known for their role in motor control, particularly in the initiation and pacing of rhythmic movement. Parkinson's disease, which produces severe deficits in rhythmic movement initiation, is caused by dopaminergic degeneration in the basal ganglia — specifically in the substantia nigra pars compacta. The same disease produces marked deficits in beat perception and musical rhythm processing. The basal ganglia are both the motor timing circuit and the musical beat-processing circuit. They are not analogous structures. They are the same structure performing the same computation in two contexts.
Grahn and Brett (2007) demonstrated in fMRI studies that beat-based rhythm — as opposed to isochronous sequences without a perceived beat — specifically activates the putamen, a basal ganglia structure directly involved in learned motor sequences. Listening to a drum pattern activates the same circuit as walking to a beat, because the same circuit is performing beat prediction in both cases.
The dopaminergic enrichment of the basal ganglia provides an additional explanatory link. Dopamine release in these circuits produces the subjective experience of reward and pleasurable anticipation. Sustained running triggers dopamine release (the "runner's high" is dopaminergic as well as endorphinergic). Musical beat entrainment also triggers dopamine release, particularly at moments of rhythmic anticipation and resolution (Salimpoor et al., 2011). The neural pleasure of music may be, in part, the evolutionary shadow of the neural pleasure of a hunt going well.
Patel (2006) proposed the Vocal Learning Hypothesis: that the capacity for voluntary rhythmic entrainment to an external beat is uniquely correlated with the capacity for vocal learning in non-human animals. Species that can entrain to a beat (certain parrots, cockatoos, elephants, sea lions) are vocal learners. Species that cannot (dogs, chimpanzees, other primates) are not. The hypothesis predicts that beat entrainment requires the tight coupling between auditory cortex and motor planning areas that is uniquely present in vocal learners — because imitating a sound requires projecting a desired acoustic output back onto motor control, which is precisely the circuit needed for adjusting movement timing to match a beat.
Humans are vocal learners of exceptional sophistication. The Persistence Entrainment Hypothesis proposes that this coupling — between hearing a rhythm and moving to it — was reinforced by the selective pressure of running, where hearing a fellow hunter's footsteps and synchronizing with them could improve energetic efficiency and group coordination. Beat perception and vocalization, in this view, co-evolved under the dual pressure of locomotor entrainment and communicative function.
The Lissajous figure for any two oscillating systems is the curve traced by plotting one oscillator's position against the other's as a function of time. When the ratio of their frequencies is rational (expressible as a simple integer fraction m:n), the Lissajous closes into a finite, repeating geometric form. When the ratio is irrational, the curve fills space chaotically and never closes.
Each gear in the Persistence Entrainment model corresponds to a distinct Lissajous figure produced by the breath-oscillator and step-oscillator at that ratio. Select a gear below to see the geometric form generated by that entrainment state. Note how the simplest ratios produce the most recognizable and elegant figures — and that the sequence of gears traces a path from simple to complex and back to simple, mirroring the composer's journey from no meter, through searching, to 4/4.
The slider below controls a sine wave oscillator from 440 Hz (Concert A, clearly a pitch) to 0.5 Hz (one cycle every two seconds, clearly a rhythm). As you move the slider left, the frequency descends through the 20 Hz perceptual boundary — the moment at which the continuous tone breaks apart into individual pulses. Above 20 Hz, you hear pitch. Below 20 Hz, you hear beats. At 2.67 Hz, you are listening to a running cadence of 160 BPM. At 1 Hz, a resting heartbeat.
The waveform display changes at the 20 Hz boundary: in pitch mode (blue), the oscilloscope shows a continuous wave. In rhythm mode (orange), it displays discrete vertical pulses — one per oscillation cycle.
The evidence assembled in this paper points toward a single conclusion: the formal properties of music — its meter, its preferred interval ratios, its preponderance of regular pulse and hierarchical rhythmic organization — are not arbitrary cultural conventions. They are the formalized traces of physiological processes that dominated the daily experience of the human animal for approximately two million years before the first instrument was fashioned, the first song was sung, or the first rhythm was deliberately composed.
The first musical composition was a run. Its time signature shifted as fatigue accumulated. Its beat was the footstrike. Its melody was the breath. Its tempo was determined by terrain, prey, and the limits of the human body at sustained aerobic maximum. It was performed by every member of the group simultaneously, in physiological unison enforced not by a conductor but by shared biology under shared physical constraint.
The drum that came later — the first deliberately constructed instrument — did not introduce rhythm into human culture. It made rhythm visible, portable, and controllable. The voice, applied to the patterns of the running breath, did not introduce melody. It rendered the pitch-trajectory of the breath — its rhythm at the sub-20 Hz level, its formant structure, its periodic compression and release — as intentional sound. The entire apparatus of music, viewed through this lens, is the externalization and voluntary reproduction of a physiological system that had been running, literally, since before the species had language.
There are predictions this hypothesis makes that are testable. If LRC entrainment is the origin of musical meter, then cultures with the strongest persistence-hunting traditions should show the most developed rhythmic (as opposed to melodic) musical practices — and the !Kung San, the Rarámuri, and the Kalahari San do. If the basal ganglia is the shared substrate for locomotor timing and musical beat, then disrupting that circuit should impair both simultaneously — and Parkinson's disease does. If the 4:4 common-time preference is a biological default rather than a cultural convention, then 4/4 (or its functional equivalent) should appear across unrelated musical traditions with no shared history of cultural contact — and it does.
Music began before language. It began before instruments. It began before the conscious intention to make music at all. It began with the body doing what bodies evolved to do: run, breathe, synchronize, and find the simplest ratio that gets all the oscillators in phase.
The rest is arrangement.