The Cochlea and Auditory Transduction

The cochlea is a coiled, fluid-filled tube in the temporal bone that performs mechanical frequency analysis on sound. Its central structural feature is the basilar membrane, which runs the length of the spiral and separates two of its three fluid-filled chambers (scala vestibuli and scala tympani; the third, scala media, sits between them). The basilar membrane is not uniform: near the base (close to the oval window) it is narrow and stiff, while near the apex it is wide and compliant, roughly a 5-fold change in width and a 100-fold change in stiffness across the ~35 mm human cochlear length. This graded mechanical property is the physical substrate for tonotopy — the orderly mapping of sound frequency onto cochlear position, high frequencies represented near the base and low frequencies near the apex. Sitting on the basilar membrane is the organ of Corti, the sensory epithelium that contains the actual mechanoreceptors of hearing: one row of inner hair cells (the primary sensory transducers, synapsing onto ~95% of afferent auditory nerve fibers) and three rows of outer hair cells (which mechanically amplify and sharpen the traveling wave rather than serving as the main sensory input). Overlying the hair bundles is the tectorial membrane, a gelatinous flap whose relative motion against the reticular lamina shears the stereocilia of the hair cells, opening mechanically gated ion channels — the transduction event proper, detailed under Hair Cell Mechanotransduction. The key conceptual move students must make is separating two stages that are often conflated: (1) cochlear mechanics — a passive-plus-active physical process that sorts frequencies by place along the basilar membrane, and (2) hair cell transduction — the conversion of local mechanical displacement into a graded receptor potential and, ultimately, neurotransmitter release onto the auditory nerve. This article is about stage one: how the cochlea builds the map that stage two merely reads out.

Step 1 — Sound enters the fluid system. Airborne sound vibrates the tympanic membrane, which drives the three middle-ear ossicles (malleus, incus, stapes). The stapes footplate pistons in and out of the oval window, injecting pressure waves into the cochlear fluids. The middle ear's lever and area-ratio action (tympanic membrane area roughly 17–20× the oval window area) provides an impedance-matching gain of about 25–30 dB, so that airborne sound energy is efficiently transferred into the much denser cochlear fluid rather than mostly reflecting off it.

Step 2 — A traveling wave forms on the basilar membrane. Pressure differences between scala vestibuli and scala tympani, communicated across the cochlear partition, deform the basilar membrane. Because the membrane's stiffness declines roughly exponentially from base to apex, a disturbance of any given frequency travels apically, grows in amplitude, reaches a peak at a frequency-specific location, then decays sharply beyond that point. High-frequency stimuli produce waves that peak near the stiff, narrow base (~20 kHz region); low-frequency stimuli travel further before peaking, cresting near the compliant, wide apex (~20 Hz–a few hundred Hz region). This is place coding by mechanical resonance and traveling-wave dynamics — not each patch of membrane "ringing" independently at its own frequency in isolation, but a coupled system whose local mechanical impedance sets where a given traveling wave stalls and peaks.

Step 3 — Local shear deflects hair bundles. At the region of peak displacement, the up-and-down motion of the basilar membrane is converted, via the geometry of the organ of Corti (the hair cells sit on a pivot point offset from the tectorial membrane's attachment), into a shearing motion between the reticular lamina and the tectorial membrane. This shear deflects the stereociliary bundles atop inner and outer hair cells. Deflection toward the tallest stereocilia increases tension on tip links, opening mechanotransduction channels; deflection the other way closes them — the actual biophysics is covered under Hair Cell Mechanotransduction.

Step 4 — Outer hair cells actively amplify (the cochlear amplifier). Passive mechanics alone predict a traveling wave far broader and less sensitive than what is actually measured in living ears. Outer hair cells solve this: depolarization triggers conformational change in the motor protein prestin embedded in their lateral membrane, causing rapid electromotility — the cell body shortens and elongates in phase with the stimulus, feeding mechanical energy back into the local basilar membrane motion. This active process amplifies small vibrations by up to 40–60 dB at the peak of the traveling wave (a ~100- to 1000-fold gain in displacement for faint sounds), sharpens the frequency tuning of the traveling wave, and — because it is an active, potentially unstable process — can produce spontaneous or evoked otoacoustic emissions, sounds generated by the cochlea itself and measurable in the ear canal.

Step 5 — Inner hair cells report the outcome to the brain. Inner hair cells, positioned in the region of peak amplified displacement, transduce local motion into receptor potentials and release glutamate onto the dendrites of auditory nerve fibers. Because each auditory nerve fiber innervates inner hair cells at essentially one cochlear location, each fiber inherits a characteristic frequency — the frequency to which it responds most sensitively at lowest threshold — directly reflecting where along the cochlea it is wired. This tonotopic order is preserved centrally, forming the basis for frequency maps throughout Auditory Pathways and Sound Localization.

The single most important body of evidence for cochlear place coding comes from Georg von Békésy, working largely in the 1930s–1950s. Using cadaveric human and animal temporal bones, Békésy built precision optical and mechanical apparatus (including a stroboscopic microscope) to directly visualize basilar membrane motion while driving the ossicular chain with pure tones of controlled frequency and intensity. He observed that:

• A displacement wave truly travels from base toward apex — it does not appear simultaneously everywhere, and it does not reflect back once past its peak.
• For any fixed frequency, the wave's envelope grows gradually, reaches one clear maximum at a location set by that frequency, and decays rapidly beyond that point (steeper decay on the apical side than the rise on the basal side).
• Higher frequencies produced envelope peaks systematically closer to the base; lower frequencies peaked systematically closer to the apex — a smooth, monotonic map, closely approximated by a logarithmic function of distance.
• Cutting or stiffening parts of the membrane, or altering cochlear fluid properties, shifted the wave's behavior in ways consistent with a mechanically graded partition, not a set of independently tuned resonant strings (an earlier hypothesis, Helmholtz's resonance theory, which Békésy's direct observations superseded).

This work — for which Békésy received the 1961 Nobel Prize in Physiology or Medicine — was the first direct physical demonstration that frequency is encoded as place along a continuous mechanical structure, rather than by separate resonating fibers as Hermann von Helmholtz had proposed in the 19th century. Békésy's cadaveric preparations, however, used relatively intense stimulation and dead (non-motile) tissue, so the tuning he observed was broader and less sensitive than tuning in the living ear. That gap was closed by later work using live, anesthetized animals and far more sensitive techniques — notably Mössbauer-effect measurements of basilar membrane velocity (pioneered by William Rhode in the 1970s using a radioactive Mössbauer source affixed to the membrane) and later laser Doppler vibrometry and laser interferometry (used extensively from the 1980s onward, e.g., by Mario Ruggero and colleagues in chinchilla cochleae). These studies showed that in a living, physiologically intact cochlea, tuning at the peak of the traveling wave is far sharper and far more sensitive (by tens of decibels) than in Békésy's cadaver preparations — and that this sharpening disappears within minutes after death or with cochlear ischemia. This demonstrated that a metabolically dependent, active process — later attributed to outer hair cell electromotility, driven by the motor protein prestin discovered by Jian Zheng, Peter Dallos and colleagues in 2000 — actively amplifies and sharpens the passive traveling wave described by Békésy. Complementary evidence comes from otoacoustic emissions, first reported by David Kemp in 1978: sounds recordable in the ear canal that are generated by the cochlea itself, which could only arise from an active, energy-producing mechanical process inside a living cochlea, exactly as predicted by cochlear amplifier theory. Together, these three converging lines of evidence — Békésy's direct optical observation of the passive traveling wave, live-animal vibrometry showing metabolically dependent sharpening, and Kemp's otoacoustic emissions demonstrating cochlear sound production — establish both halves of the modern mechanism: a passive, physically graded traveling wave that sorts frequency by place, actively amplified and sharpened by outer hair cells.

Example 1 — Reading the tonotopic map (place-frequency conversion). The place-frequency relationship along the human basilar membrane is often approximated by the Greenwood function: f = A(10^(a·x) − k) where x is fractional distance from the apex (0 to 1), and for humans approximately A ≈ 165.4, a ≈ 2.1, k ≈ 0.88 (frequency in Hz). Suppose a clinical electrode array for a cochlear implant is inserted to a point that is 70% of the distance from the apex toward the base (x = 0.70). What characteristic frequency does that location normally represent? Step 1: compute the exponent term: a·x = 2.1 × 0.70 = 1.47, so 10^1.47 ≈ 29.5. Step 2: subtract k: 29.5 − 0.88 ≈ 28.6. Step 3: multiply by A: 165.4 × 28.6 ≈ 4,730 Hz. So an electrode contact sitting at 70% of the cochlear length from the apex lies in a region that, acoustically, would normally represent roughly 4.7 kHz — solidly in the range important for consonant sounds in speech. This calculation is exactly why cochlear implant programming must map electrode position to frequency channel: stimulating the wrong place produces a systematically mispitched percept, and implant users often need weeks of auditory training to reinterpret a compressed, shifted tonotopic map as normal speech.

Example 2 — Middle-ear gain and a hearing-loss scenario. A patient presents with conductive hearing loss from otosclerosis (stapes footplate fixation) that reduces effective middle-ear transmission by 30 dB, while their cochlea and hair cells remain entirely normal. A clinician wants to know: at what sound pressure level (SPL) must a 1 kHz tone be presented in air for this patient to receive the same cochlear input as a normal-hearing person hears at a comfortable conversational level of 60 dB SPL? Reasoning: normal middle-ear transduction delivers roughly 60 dB SPL worth of cochlear stimulation for a 60 dB SPL airborne tone (by definition of SPL calibration in normal ears). With a 30 dB conductive loss, the same 60 dB airborne tone now delivers only the cochlear equivalent of a 30 dB SPL signal — likely below threshold for many normal-hearing listeners and certainly too faint for comfortable perception. To restore the same 60 dB-equivalent cochlear drive, the external tone must be raised by the amount of the conductive loss: Required external SPL = target cochlear-equivalent level + conductive loss = 60 dB + 30 dB = 90 dB SPL. This is precisely the logic behind hearing-aid gain prescription for conductive loss: because the cochlea itself is intact, simply amplifying the incoming sound by an amount equal to the conductive deficit (here 30 dB) restores normal-equivalent cochlear drive, and speech discrimination is typically excellent once loudness is restored — unlike sensorineural loss from hair cell damage, where amplification alone cannot restore fine frequency resolution because the active cochlear amplifier itself is degraded.

Cochlear implants exploit tonotopy directly: an electrode array threaded into the scala tympani delivers electrical pulses to different cochlear locations, with basal electrodes carrying high-frequency channel information and apical electrodes carrying low-frequency information, deliberately mimicking the natural place code even though hair cells are bypassed entirely.

Hearing-aid fitting and audiometry rely on the frequency-place map to interpret pure-tone audiograms: a notch at 4 kHz on an audiogram (the classic "noise notch") reflects damage concentrated at the basal cochlear region tuned to that frequency, consistent with the physical vulnerability of that high-stiffness, high-velocity region to acoustic trauma.

Otoacoustic emission testing is now standard in newborn hearing screening worldwide: because emissions require living, functioning outer hair cells, their absence flags cochlear dysfunction within days of birth, enabling early intervention during the critical period for language development.

Ototoxicity monitoring in oncology and infectious-disease treatment (e.g., cisplatin, aminoglycoside antibiotics) uses knowledge of tonotopy to predict and track hearing loss, since these drugs preferentially damage basal, high-frequency-coding outer hair cells first, producing a predictable audiometric signature clinicians watch for.

Tinnitus and hyperacusis research increasingly implicates disordered cochlear amplifier function and subsequent central reorganization of the tonotopic map, linking cochlear mechanics directly to a widespread and still poorly treated clinical condition, and motivating ongoing work connecting cochlear place coding to central plasticity described under Auditory Pathways and Sound Localization.

Ask the tutor: How does damage to outer versus inner hair cells produce different patterns of hearing loss on a clinical audiogram?