3How Bone Conduction Works
Shake the skull and both cochleae hear. The physics of how vibration reaches the inner ear explains why bone conduction bypasses a diseased middle ear and why it never quite belongs to one side.
FThe skull as a sound delivery system
Bone conduction begins with a simple fact: the cochlea is encased in the densest bone in the body, the otic capsule of the temporal bone, and that bone is part of the skull. Vibrate the skull and you vibrate the cochlea with it. A vibrating actuator pressed against or fixed to the head sets the whole skull oscillating, and because both cochleae are embedded in that same vibrating shell, both inner ears are stimulated, regardless of which side the actuator sits on. This is utterly unlike air conduction, where a sound enters one ear canal and chiefly drives one cochlea.
A century of experiments confirms that bone-conducted and air-conducted sound end up exciting the cochlea in the same way. In von Bekesy’s classic demonstration, a tone delivered by bone and the same tone delivered by air could be made to cancel each other by adjusting phase and amplitude, proving that both ultimately drive the same travelling wave along the basilar membrane. Whatever path the energy takes through the head, the final common event is the motion of the cochlear partition and the bending of hair cells.[2015][2021]
TThe pathways that turn skull motion into hearing
Skull vibration does not reach the basilar membrane by a single mechanism. Tonndorff catalogued several distinct contributions, and later systematic work by Stenfelt and Goode sorted out which dominate at which frequencies. One family is inertial: when the skull accelerates, the middle-ear ossicles lag behind because of their mass, producing relative motion at the stapes, and the cochlear fluids likewise lag, producing a pressure difference across the partition. A second family is compressional: at high frequencies the otic capsule itself is alternately squeezed and released, distorting the cochlear ducts directly.
The frequency split is clinically useful to remember. Below roughly one kilohertz the inertia of the cochlear fluid is the leading contributor. Through the mid frequencies the inertia of the middle-ear contents takes a larger share. Above about four kilohertz, compression of the otic capsule dominates. Because several of these pathways do not depend on an intact, mobile ossicular chain or a patent ear canal, bone conduction continues to work even when the outer and middle ear are diseased, which is precisely the property that makes it therapeutically valuable.[2024][2021]
TTranscranial attenuation: why both ears, but not equally
Although a single actuator drives both cochleae, it does not drive them identically. As vibration travels from the stimulation site across the head to the far cochlea it is attenuated and delayed, and this loss is called transcranial attenuation. It is famously quoted as roughly ten decibels, but that figure hides large variability: it changes with frequency and differs substantially between individuals, with standard deviations of several decibels in classic measurements. The head is not a uniform sphere; intracranial contents, sutures and geometry create resonances and sharp antiresonant dips where vibration can drop by many decibels over a narrow frequency band.
This near-symmetry has two clinical faces. It is an asset in single-sided deafness, where a device on the deaf side can shuttle sound across the skull to the one working cochlea, lifting the head-shadow effect. It is a liability in bilateral fittings, because the limited separation between the two cochleae blurs the interaural differences the brain uses to localise sound, so two bone devices do not deliver the clean binaural separation that two air-conduction aids can. Understanding transcranial attenuation is therefore central to predicting what a bone device can and cannot achieve.[2020][2024]
CFrom physics to coupling
The physics dictates how a device should be coupled to the head. Because intact skin damps vibration, a percutaneous abutment that drives bare bone delivers more output, especially in the high frequencies, than a transducer that must vibrate through skin. An active transcutaneous implant solves the same problem differently by placing the powered transducer under the skin directly on the bone, so only the signal crosses the skin. In every case the goal is a rigid, low-loss mechanical link between the actuator and the skull, because any compliant interface bleeds away the vibration the cochlea needs.
The coupling site matters less than one might fear, since the whole skull moves, but it is not irrelevant: placement closer to the target cochlea and a stiffer fixation improve output, and the inevitable cross-talk to the contralateral cochlea must be anticipated when planning bilateral or single-sided fittings. The same physics also limits bone devices. Output is capped by feedback from sound radiating off the vibrating head and by the energy needed to move the entire skull, which is why bone devices excel at closing an air-bone gap but struggle to lift thresholds far beyond the patient’s own bone-conduction curve.[2015][2024]
What is the correct physical explanation?
Von Bekesy's cancellation experiment demonstrated that bone-conducted and air-conducted sound:
Below about 1 kHz, the dominant bone-conduction pathway is:
Why does bone conduction keep working when the outer and middle ear are diseased?
The often-quoted transcranial attenuation of about 10 dB is best described as:
For device coupling, the physics implies that the ideal interface to the skull is: