The Nervous System · Part Two · How It Senses and Moves

23Sensory

Lesson 23 / 61

Sensory Processing: Labeled Lines, Rate Codes, and Descending Control

Which cell fires, how fast it fires, and who edits the stream.

Sensory processing decides what an arriving nerve impulse means. Which receptor cell fires, and which pathway carries it, sets the modality. Discharge rate carries intensity, and the relative timing of the first spikes across an ensemble carries what rate cannot. None of that traffic arrives unedited. Descending fibers reach the thalamic relay, the hair cells of the inner ear, and the muscle spindle. The Unified Model of Tone treats the settings those fibers hold as the reason one nervous system reads a volley loud where another reads it quiet.

Labeled lines

Mice without PKD2L1 taste cells lost sour alone; every other tastant unaffected

Flutter band

Skin vibration is felt as flutter from 5 to 50 Hz

Duty cycle

A 0.5 ms spike at 50 per second leaves the neuron silent 97.5% of the time

Convergent streams

Visual, vestibular, auditory, somatosensory, visceral

Sensory processing

A receptor turns a physical event into membrane current, then into a train of action potentials, and every train crosses at least one subcortical nucleus before it reaches cortex. Four low-threshold mechanoreceptive types share one patch of glabrous skin in the human hand: fast-adapting, Pacinian, SA-I, and SA-II units. Each answers a different feature of the same touch, so one contact is reported four ways at once.

Multisensory convergence

Single cells receive more than one sense. In rhesus monkey superior colliculus, deep layer neurons answering two modalities produced responses far exceeding the sum of the two alone, and the gain was largest when each single input was weakest. The arithmetic runs both ways. Two stimuli in different places produced no interaction or outright depression of the response, so what a convergent cell pays for is agreement between its streams.

01Labeled lines in skin

Labeled lines set the modality of a sensation by which cell fires

The labeled line principle states that information from a particular receptor travels a specific pathway to a specific part of the nervous system. Modality depends on which cell, pathway, nucleus, or lobe the stimulus activates. Genetic subtraction proves it. Mice engineered without PKD2L1-expressing taste cells were completely devoid of sour responses, and every other tastant was unaffected Huang 2006. Delete one population of receptor cells and exactly one modality leaves with it.

A receptor does not transmit warmth or pressure as such. It transmits action potentials, and the destination decides the meaning. Touch reaches the somatosensory cortex, sound reaches the auditory cortex, and the difference is wiring rather than waveform.

The address beats the molecule in the nematode too. In C. elegans, AWC neurons respond to low odorant concentrations and ASH neurons to high ones, so the same odorant attracts at low concentration and repels at high Yoshida 2012. The authors read odor concentration coding there as mostly conforming to the labeled line principle. One molecule, two cells, two opposite worlds.

The map lays down more lines where it needs resolution

Human glabrous skin carries four labeled types at once. Microneurography sampled 334 units in all. Innervation density was estimated at 241 units per square centimeter at the fingertip against 58 in the palm Johansson 1979. Relative densities run 1 in the palm, 1.6 in the finger, and 4.2 at the tip. The absolute figures come from a model built on histological fiber counts rather than from a direct count.

The types differ in hardware, not only in name. Of 101 slowly adapting receptors recorded in human glabrous skin, 88 were SA-I and 13 were SA-II Knibestöl 1975. SA-I fibers conducted at 58.7 +/- 2.3 m/s and SA-II fibers at 45.3 +/- 3.6 m/s. SA-I dynamic thresholds ran 0.15 to 1.35 mm and SA-II thresholds 0.25 to 0.95 mm.

The two types also idle differently. SA-I receptors showed no spontaneous discharge and a mostly irregular pattern, while most SA-II receptors discharged spontaneously and very regularly Knibestöl 1975. SA-II units responded to lateral stretching of the skin and often showed directional sensitivity, which is how the two classes were separated. Static thresholds separate them again, running from 0.25 mm to more than 2.0 mm for SA-I and from 0.55 to 1.65 mm for SA-II. A labeled line carries a resting habit as well as a modality.

Clinicians read the principle in reverse. Because identity rides on the line, the modality a patient can no longer register localizes the damaged cell, tract, or nucleus. Labeled lines turn the dense interior of the nervous system into named channels, and the silence of a channel has an anatomical address.

02Findings

What the research shows

241 against 58
Estimated mechanoreceptive unit density per square centimeter at the human fingertip against the palm, from 334 sampled units and a model built on histological fiber counts Johansson 1979. Resolution is laid down anatomically, so the map of labeled lines is denser where the hand needs detail.
0.7 against 1.0
Exponents of the neural and psychophysical stimulus-response functions recorded from the same human subjects at the same moment Knibestöl 1980. The neural function decelerated across the force range while the psychophysical function ran linear.
8 micrometers
Cupula deflection amplitude below which afferent spike rate did not change in larval zebrafish, whatever the velocity Haehnel-Taguchi 2014. Above that floor, maximum spike rate rose linearly with deflection velocity between 0.1 and 1.0 micrometers per millisecond. A labeled line has a floor, and beneath it the world is unavailable to that line.
2 dB to 30 dB
Middle ear attenuation below 1 kHz near hearing threshold against high sound levels, with an attack time about 100 ms, in three experienced human listeners Sesterhenn 1978. Gain control sits ahead of the receptor, which is why cochlear excitation holds near one value from threshold to well above 70 dB HL.
30 up, 33 down
Guinea pig medial geniculate neurons potentiated by 8.6 +/- 5.5 mV against those hyperpolarized by 11.3 +/- 4.9 mV when auditory cortex was activated, out of 63 modulated cells Yu 2004. Descending control routes the afferent stream rather than muting it.
27.8 percent
Share of deep layer superior colliculus neurons in rhesus monkey responding to more than one sensory modality, with interactions inside a 500 ms window Wallace 1996. Convergence is built into single cells, so no stream is ever read alone.
13 against 10
Neck pain patients and healthy controls given identical 100 Hz neck vibration Beinert 2015. Joint position sense improved in the patient group and acuity fell in the control group, so the stimulus did not carry the outcome and the starting state did.
20 minutes
Duration of the N20 and N30 amplitude decrease after one session of cervical spine manipulation in 12 subjects, against no change in 12 passive head movement controls Haavik-Taylor 2007. Input at the neck reached the cortical processing of an input from the arm.

03Rate coding of intensity

The frequency code reports intensity as spikes per second, and perception does not read it straight off

Mammalian auditory nerve fibers compress a 40 dB span of sound pressure, a hundredfold range, into firing rates climbing to about 300 Hz Ospeck 2012. An action potential is all or none, so a firmer press cannot make a larger spike. It makes a denser train, and mechanoreceptors code the intensity of a mechanical stimulus by modulating that rate.

Discharge rate is not felt magnitude. Microneurography recorded slowly adapting afferents in the human hand while the same subjects rated how strong the touch felt. The neural stimulus-response function fitted an exponent of 0.7 and the psychophysical function an exponent of 1.0, with neural exponents spanning 0.26 to 1.92 and psychophysical exponents 0.36 to 2.09 Knibestöl 1980. Nerve and person were measured at the same instant and did not agree.

The Unified Model of Tone takes that gap as the substance of sensory processing. An exponent of 0.7 at the nerve describes a decelerating function, in which each added gram of force buys less firing than the gram before it. The same subjects rated the same series as a straight line. Something between the axon and the report multiplies the upper part of the range back up. Tone shows up here as that multiplication, held in the receiving system rather than at the ending. The authors reached the same door from the other side, concluding that the shape of the psychophysical function depends heavily on central mechanisms.

Rate carries the decision and timing carries what rate cannot

Flutter is felt when the skin vibrates between 5 and 50 Hz. In awake monkeys discriminating flutter, spike train periodicity ran extremely high in S1 and was almost absent in S2. Periodicity did not covary with performance on single trials, while firing rate modulations correlated significantly with psychophysical performance Salinas 2000. Rate carried the perceptual decision at that level, and the authors leave open whether timed spikes contribute elsewhere.

The gaps still do work. An action potential lasting 0.5 milliseconds, repeating at 50 per second, leaves the neuron silent for 97.5 percent of the time. In some cells, such as pyramidal tract neurons, a particular interval between potentials may signal the start or stop of a movement. In human tactile afferents, the relative timing of the first impulses across an ensemble conveys the direction of fingertip force and the shape of the contacting surface Johansson 2004. It arrives faster than any rate code could deliver it.

Above the single cell, pattern codes represent stimulus features in the activity of whole populations, decoded by comparison, addition, or subtraction. In the grasshopper auditory pathway, a three-layer feed-forward network of a few dozen neurons converts a summed population code at the input into a labeled line population code Clemens 2011. Two processing steps carry the whole transformation. Summed coding and labeled coding are stages of one operation, not rival schemes.

04Lateral line hair cells

The fish lateral line runs the same hair cell transducer the human inner ear runs

The fish lateral line and the human inner ear run the same transducer: a hair cell in a high potassium bath, deflected by micrometers. The lateral line detects water movement along the flanks of the body and drives orienting and swimming reflexes, working through clusters of hair cells called neuromasts, each capped by a gelatinous cupula. Mechanical disturbances near the animal create pressure waves that deflect the cupula and bend the hair cells beneath it.

Juvenile blind cavefish, Phreatichthys andruzzii, carry 480 to 538 superficial neuromasts each, 362 to 410 of them on the head and 116 to 140 on the trunk. Each hair cell is pear-shaped, 15 to 21 micrometers high and 4 to 6 micrometers in diameter, with a single long kinocilium and several short stereocilia Dezfuli 2009. Those counts come from five scanning-microscopy specimens of one blind species at six months old.

Deflect one neuromast and the connected afferent reports velocity as rate. In larval zebrafish, maximum afferent spike rate increased sigmoidally with cupula deflection velocity, with a linear range between 0.1 and 1.0 micrometers per millisecond Haehnel-Taguchi 2014. Below 8 micrometers of deflection amplitude the spike rate did not change at all, whatever the velocity. Stimulating a single neuromast could elicit a swimming response, so the arc from cupula to motor output was measured end to end.

Hair cells in fish and humans work in the same high potassium bath

That bath carries a standing charge, and the charge is not the same in every organ. The vestibular endolymphatic potential runs +1 to +11 mV, with reports of up to +70 mV in the cupula of the semicircular canal ampullae. Across vertebrate groups the auditory endolymphatic potential climbs from near 0 mV in non-mammals to roughly +100 mV in the mammalian cochlea Köppl 2018. The proposed selective pressure for that escalation is holding the alternating component of the hair cell receptor potential steady at high frequencies.

The kinship has a node in time. A review of gene networks across lancelets, tunicates, and vertebrates places the hair cell bearing inner ear and lateral line as major evolutionary advances specific to vertebrates Fritzsch 2024. It calls homology to pre-vertebrate mechanosensory cells questionable, on the expression of the transcription factors Neurog and Atoh1/7. The inner ear is a lateral line turned inward and roofed over, and the transducer it reuses is a vertebrate invention.

Reading the lineage this way reframes human balance as ancient sensing. The cupula of the fish and the cupula of the semicircular canals share more than a name. They share a scale, deflections measured in single micrometers, and that scale is the layer vestibular and proprioceptive traffic works in. To examine balance is to examine a system older than limbs.

05Descending fibers at receptors

Descending fibers route the stream at the relay and at the epithelium

Sensory input is edited on the way in. The receptors of the inner ear and the muscle spindles sit under descending influence, which protects them from functional damage by strong stimuli and from adaptation to constant ones. Middle ear attenuation is level dependent. Between hearing threshold and 70 dB HL it costs only about 2 dB below 1 kHz, and at higher levels it reaches 30 dB, with an attack time near 100 ms Sesterhenn 1978. Excitation of the cochlea stays nearly constant across that range, measured in three experienced human listeners.

Remove the regulator and the arrangement shows itself. A stapes fixed by otosclerosis produces conductive recruitment, because the middle ear can no longer trade sensitivity against level Sesterhenn 1978.

In the echolocating bat Myotis lucifugus, the middle ear muscles contract synchronously with the animal's own loud vocalization, on a command issued by the vocalization system. The acoustic reflex cannot do that job. Its shortest latency was 3 to 4 msec by electromyogram and 4 to 8 msec by attenuation of the cochlear microphonic Suga 1975. The muscles are already contracting before the pulse leaves the larynx. The contraction is timed by the vocalization command, not triggered by the sound. The stapedius of that bat fused into tetanus at 260 to 320 contractions per second, and the olivocochlear bundle showed no sign of attenuating self-stimulation in the same preparation.

Attention edits the stream at the relay and at the receptor

All sensory signals pass through several subcortical nuclei before reaching cortex, and those relays are where descending influence does its filtering. Attention modified electrical activity in the cochlear nucleus of unanesthetized cats in the first demonstration of the effect Hernandez-Peon 1956. Intracellular recording supplied the mechanism half a century later.

Electrical activation of auditory cortex modulated the membrane potential of 63 guinea pig medial geniculate neurons. Thirty were potentiated by 8.6 +/- 5.5 mV and 33 were hyperpolarized by 11.3 +/- 4.9 mV Yu 2004. The potentiations lasted 125.5 msec and the hyperpolarizations 210.0 msec. Nine of the cells were recovered histologically, and in those the split matched facilitation of lemniscal relay cells against inhibition of non-lemniscal ones. Descending traffic performs selective routing rather than a change of volume.

In people the edit reaches the sensory epithelium. Efferent suppression of cochlear output was larger during the memory period of a visual working memory task in 21 normal-hearing adults than in 10 passively exposed controls Marcenaro 2021. The load was four visual objects against two, the assay was contralateral suppression of distortion product otoacoustic emissions, and the sounds being suppressed were distractors the task did not need. Holding a visual load in mind changed the gain of the ear.

The hardware is old. The alpha9 and alpha10 nicotinic subunits build the receptor through which efferent fibers inhibit hair cells. They are expressed in the mechanosensory hair cells of the vertebrate inner ear and the lateral line alike Lipovsek 2021. The authors propose a link between the evolution of the octavolateral system and the divergence of these receptors. Efferent control shipped with the hair cell.

The muscle spindle is edited the same way, by fusimotor fibers that set the receptor's sensitivity before the stretch arrives, and the human recordings sit in muscle spindles and proprioception.

06Convergence and reference frames

Convergence gives single cells several senses, and the frames those cells merge do not match

Single cells in the superior colliculus and the parietal cortex answer to more than one sense at once. Multisensory signals from the visual, vestibular, auditory, somatosensory or proprioceptive, and visceral receptors act on many sites in the central nervous system in parallel. They supply cognitive and emotional reference to the cerebral cortex and the limbic system, so a body may answer the same input with a voluntary movement or with emotional motor behavior depending on the context.

The superior colliculus supplies the census. In rhesus monkey, 27.8 percent of deep layer neurons responded to stimuli from more than a single sensory modality, against 37 percent visual only, 17.6 percent auditory, and 17.6 percent somatosensory Wallace 1996. The modality maps lie in spatial register, and interactions appear when two stimuli fall within 500 ms of each other, with products far exceeding the sum of the unimodal responses.

Two rules govern the arithmetic. Inverse effectiveness holds that the weaker each single input, the larger the multisensory gain. Spatial disparity works the other way, and stimuli in different places produce no interaction or outright response depression Wallace 1996. Convergence pays for agreement between streams and charges for conflict.

In the Unified Model of Tone the second rule carries more weight than the first. A body holding a distorted report from one stream loses more than that stream. It puts the remaining streams into disagreement at cells that were computing well before, and the cost shows up as depression at neurons no lesion has touched.

Sensory convergence does not remap every modality into one frame

In macaque ventral intraparietal area, tactile receptive fields were head-centered while visual fields sat between eye-centered and head-centered coordinates Avillac 2005. That area receives converging input from the visual, somatosensory, auditory, and vestibular systems, and those systems use diverse reference frames to encode sensory information. The findings are inconsistent with a remapping of all modalities into a common frame, and support integration by multidirectional sensory prediction across frames. Convergence in sensory processing is a negotiation between coordinate systems that never fully agree.

Postural control shows what the negotiation is for. Automatic regulation of postural muscle tone and basic reflexes runs through the brainstem and spinal cord, while cognitive postural control becomes decisive when a person learns a new motor skill or enters unfamiliar terrain. Because every modality converges, a change in one stream reweights the whole, and the reweighting is what a body stands on.

07Neck input and perception

Input from the cervical spine changes what the rest of the sensory system reports

Cervical proprioceptive input sits inside the computation that decides which way is up. Vibrating the dorsal neck muscles of 26 healthy adults shifted visually perceived roll McKenna 2004. Subjects were tested upright and at 30 degrees of static left and right head roll-tilt. In the tilted conditions, with vibration applied opposite the tilt, upper cervical vibration shifted perception more than lower cervical or mastoid vibration.

The most striking version comes from three patients with right hemisphere lesions and no visual field defects. Neglect of stimuli in the contralesional left visual field was reduced by vibrating the left posterior neck muscles, and reduced again by turning the trunk 15 degrees to the left Karnath 1993. Vibrating hand muscles did nothing. Turning the head rather than the trunk did nothing. The effect belonged to the head-on-trunk signal in a case series of three, and the authors offer their mechanism as a hypothesis.

Neck vibration moves patients and healthy controls in opposite directions

Short-term 100 Hz vibration of the neck muscles improved cervical joint position sense at p below .01 and reduced dynamic postural sway at p below .05 in 13 patients with neck pain. The identical stimulus reduced joint position sense acuity at p below .05 in 10 healthy controls, alongside a nonsignificant increase in sway Beinert 2015. One frequency, one tissue, opposite signs in 13 patients and 10 controls.

The Unified Model of Tone reads that split as the receiving system doing the deciding. A neck whose report has drifted is being scored wrongly upstream, and 100 Hz sharpens a report the system has been mis-reading. A neck whose report is already accurate has nothing to correct, and the same train adds noise to a channel that was scoring well. That predicts an arithmetic rather than a direction. Joint position sense error in degrees should fall furthest in the subjects who started furthest out, barely move in those who started accurate, and the spread across the group should narrow. A stimulus whose effect belonged to the stimulus would move accurate and inaccurate subjects the same way and leave the variance where it was. Beinert's two groups moved in opposite signs from one frequency at one tissue, which is the result the second account does not produce.

The cortex registers segmental input

A single session of cervical spine manipulation decreased the amplitude of the parietal N20 and frontal N30 somatosensory evoked potential components Haavik-Taylor 2007. The 12 subjects all had a history of recurring neck stiffness or pain, and the change lasted an average of 20 minutes. Twelve separate controls given passive head movement showed no change. The evoked potentials were driven by median nerve stimulation at the wrist, so input delivered at the neck changed how the cortex read traffic arriving from the arm.

Where the thrust lands decides whether the drop appears at all. The drop follows a thrust at a segment chosen on clinical indicators, and a thrust at a predetermined segment produces no significant change Niazi 2024. The trial randomized 96 adults with recurrent mild neck pain, double-blinded and active-controlled. The research page on tone carries that trial in full. Go there for what the N30 drop means about analysis and specificity. Stay here for the labeled line and the descending wiring that carries it.

Weighting is what an input changes, and the cortex reading a spinal segment is the same cortex reading the hand. A chiropractic adjustment is graded mechanical and sensory information delivered at one segment of that shared map. It arrives in a stream descending fibers are already editing at the relay and at the receptor, so the cortical reading reports the input and the editing together.

That is why one weighting can be read through instruments that share no units. Amplitude in microvolts at the N20 and N30 comes from the scalp. Joint position sense error comes in degrees at the neck. Sway comes from a force platform under the feet. Each instrument samples the same organization at a different port, which is what a chord and not a note means at this level of the nervous system.

A touch is felt at the magnitude the nervous system has already decided to give it.

08Tone

How this system expresses tone

Tone shows up here as the settings the system holds before any traffic arrives. Descending fibers reach the middle ear muscles, the hair cells of the cochlea, and the intrafusal fibers of the spindle. What they hold there decides how loud a channel arrives and which channels the system agrees to read together.

Input quality

Fidelity is fixed at the ending. SA-I units in human glabrous skin answered dynamic indentations of 0.15 to 1.35 mm, and a movement under that threshold is never reported.

Gain

Gain is set ahead of the input. The bat contracts its middle ear muscles on the vocalization command, since the reflex latency of 3 to 4 msec comes too late.

Coupling

Streams are read against each other. Vibrating the left posterior neck muscles reduced left-sided neglect in three patients with right hemisphere lesions, while vibrating hand muscles did nothing.

Oscillation is flutter between 5 and 50 Hz, carried as a periodic spike train that runs strong in S1 and is almost gone in S2. Load is the 0.5 ms spike at 50 per second that leaves a neuron silent 97.5 percent of the time, which is what makes a rate code affordable. Prediction is macaque area VIP holding tactile fields head-centered while visual fields sit between eye-centered and head-centered coordinates. Set point is the middle ear trading sensitivity against level, spending about 2 dB below 1 kHz near threshold and up to 30 dB at high levels. Cochlear excitation therefore holds near one value from hearing threshold to well above 70 dB HL. Time course is a middle ear attack of about 100 ms sitting inside a cortical change that lasted 20 minutes after one cervical input. Constraint is the 8 micrometers of cupula deflection below which lateral line afferents in larval zebrafish do not change their rate. Part of the physical world is unavailable to that system by construction.

09Across the library

How this page relates to the rest of the library

The senses and the nervous system

A threshold read as an adjustable setting rather than a fixed property of the ear or the eye, and what a percept assembled from several channels costs when one channel is scored wrongly. Go there for what a threshold means and what moves it. Stay here for the labeled line, the rate code, and the efferent wiring that reaches the receptor.

Proprioception and Interoception: What the Body Knows About Itself

The body's own report about itself, the five ways that report fails, and what a failed report costs the regulation running on it. Go there for what a degraded report costs downstream. Stay here for the receptor cells that write the report and the codes they write it in.

Muscle spindles and proprioception

Inside the capsule whose afferents fire at a mean 11.1 Hz at rest. The intrafusal fibers and the fusimotor loop live there, along with the human recordings that read the loop from both the afferent and the efferent side.

The peripheral nerve and reflex

The cable that carries these labeled lines, including the fiber classes behind the 58.7 m/s of SA-I and the 45.3 m/s of SA-II, and the segmental loop the afferents close.

Cranial nerves VII to XII

Where the octavolateral inheritance becomes the eighth nerve, carrying hearing and equilibrium into the brainstem from the same hair cells the lateral line runs on.

Sensory integration in children

What convergence looks like while the weighting is still being built, when the streams have not yet settled on how much to trust each other.

The vestibular system

Where the cupula of the fish becomes the cupula of the semicircular canals, and what the balance system does with hair cell traffic once the eighth nerve delivers it.

10Frequently asked

Questions about this topic

What is sensory processing?

Sensory processing is the work by which receptors convert physical events into nerve traffic the brain reads as modality, intensity, and timing. A receptor cell turns pressure, sound, chemistry, or body position into membrane current, then into action potentials. Those trains cross at least one subcortical nucleus before reaching cortex, and descending fibers edit them on the way. Four low-threshold mechanoreceptive types share the glabrous skin of the hand, at densities estimated as high as 241 units per square centimeter at the fingertip.

What is the labeled line principle?

The labeled line principle states that information from a particular receptor travels a specific pathway to a specific part of the nervous system. The modality of a sensation therefore depends on which cell, pathway, nucleus, or lobe the stimulus activates. Genetic ablation demonstrates it. Mice lacking PKD2L1-expressing taste cells were completely devoid of sour responses and answered every other tastant normally. The same logic runs in the nematode, where one class of neuron reports low odorant concentration and another reports high concentration of the identical molecule.

How does the nervous system code how strong a stimulus is?

Intensity is coded as discharge rate. An action potential is all or none, so a firmer press produces a denser train rather than a bigger spike. Mammalian auditory nerve fibers compress a 40 dB span of sound pressure, a hundredfold range, into firing rates reaching about 300 Hz. Rate is not the whole code. The relative timing of the first spikes across an ensemble of human tactile afferents carries the direction of fingertip force and the shape of the contacting surface, faster than any rate code could deliver it.

Can the brain change what the ear and the muscles send it?

Descending control reaches all the way to the receptor organ. The middle ear muscles remove about 2 dB below 1 kHz near hearing threshold and up to 30 dB at high sound levels, with an attack time near 100 ms. Excitation of the cochlea therefore stays nearly constant. Efferent suppression of cochlear output increased in 21 adults holding a visual working memory load. In the thalamus, activating auditory cortex potentiated 30 guinea pig relay neurons and hyperpolarized 33 others, which makes descending traffic a routing decision.

What does the fish lateral line have to do with human balance?

The lateral line detects water movement through neuromasts, clusters of hair cells capped by a gelatinous cupula. Juvenile blind cavefish carry 480 to 538 of them, each hair cell 15 to 21 micrometers high with one kinocilium and several stereocilia. The vertebrate inner ear runs the same transducer in the same high potassium bath. The endolymphatic potential escalates from 1 to 11 mV in the vestibular system to roughly 100 mV in the mammalian cochlea. A gene network review places both systems at the vertebrate node.

Why does trouble in one sensory stream cost more than that stream?

A drifted stream does not fail alone, because convergent cells read it against the others. In rhesus monkey superior colliculus, 27.8 percent of deep layer neurons answered more than one modality, and two stimuli falling within 500 milliseconds produced responses far exceeding the sum of the single-modality responses. Stimuli in different places depressed those same cells instead. A stream reporting the wrong place therefore puts the remaining streams into disagreement, and the cost appears as depression at neurons no lesion has touched.

Why can neck problems change balance or vision?

Cervical proprioceptive input sits inside the computation that decides which way is up. Vibrating the dorsal neck muscles of 26 healthy adults shifted visually perceived roll, and upper cervical vibration shifted it more than lower cervical or mastoid vibration when the head was tilted 30 degrees. In three patients with right hemisphere lesions, vibrating the left posterior neck muscles reduced left-sided neglect, while vibrating hand muscles did nothing. The neck reports where the head sits on the trunk, and vision is read inside that frame.

Why does the same sensory input help one person and not another?

The outcome is set by the system receiving the input. Short-term 100 Hz vibration of the neck muscles improved joint position sense and reduced dynamic postural sway in 13 patients with neck pain. The identical stimulus reduced joint position sense acuity in 10 healthy controls. One frequency, one tissue, opposite signs. The Unified Model of Tone reads the split as a difference in what there was to correct: a neck already reporting accurately has nothing to gain from the input and a well-scored channel to disturb.

11The sources

References

1
Huang AL, et al.. The cells and logic for mammalian sour taste detection. Nature. 2006. PMID 16929298
2
Yoshida K, et al.. Odour concentration-dependent olfactory preference change in C. elegans. Nat Commun. 2012. PMID 22415830
3
Johansson RS, Vallbo AB. Tactile sensibility in the human hand: relative and absolute densities of four types of mechanoreceptive units in glabrous skin. J Physiol. 1979. PMID 439026
4
Knibestöl M. Stimulus-response functions of slowly adapting mechanoreceptors in the human glabrous skin area. J Physiol. 1975. PMID 1127614
5
Ospeck M. Fast negative feedback enables mammalian auditory nerve fibers to encode a wide dynamic range of sound intensities. PLoS One. 2012. PMID 22412868
6
Knibestöl M, Vallbo AB. Intensity of sensation related to activity of slowly adapting mechanoreceptive units in the human hand. J Physiol. 1980. PMID 7381785
7
Salinas E, et al.. Periodicity and firing rate as candidate neural codes for the frequency of vibrotactile stimuli. J Neurosci. 2000. PMID 10884334
8
Johansson RS, Birznieks I. First spikes in ensembles of human tactile afferents code complex spatial fingertip events. Nat Neurosci. 2004. PMID 14730306
9
Clemens J, et al.. Efficient transformation of an auditory population code in a small sensory system. Proc Natl Acad Sci U S A. 2011. PMID 21825132
10
Dezfuli BS, et al.. Ultrastructure and distribution of superficial neuromasts of blind cavefish, Phreatichthys andruzzii, juveniles. Microsc Res Tech. 2009. PMID 19343789
11
Haehnel-Taguchi M, et al.. Afferent and motoneuron activity in response to single neuromast stimulation in the posterior lateral line of larval zebrafish. J Neurophysiol. 2014. PMID 24966296
12
Köppl C, Wilms V, Russell IJ, Nothwang HG. Evolution of Endolymph Secretion and Endolymphatic Potential Generation in the Vertebrate Inner Ear. Brain Behav Evol. 2018. PMID 30415265
13
Fritzsch B, et al.. Gene networks and the evolution of olfactory organs, eyes, hair cells and motoneurons: a view encompassing lancelets, tunicates and vertebrates. Front Cell Dev Biol. 2024. PMID 38533086
14
Sesterhenn G, Breuninger H. On the influence of the middle ear muscles upon changes in sound transmission. Arch Otorhinolaryngol. 1978. PMID 697648
15
Suga N, Jen PH. Peripheral control of acoustic signals in the auditory system of echolocating bats. J Exp Biol. 1975. PMID 1206335
16
Hernandez-Peon R, Scherrer H, Jouvet M. Modification of electric activity in cochlear nucleus during attention in unanesthetized cats. Science. 1956. PMID 13298689
17
Yu YQ, et al.. Corticofugal gating of auditory information in the thalamus: an in vivo intracellular recording study. J Neurosci. 2004. PMID 15044545
18
Marcenaro B, et al.. The medial olivocochlear reflex strength is modulated during a visual working memory task. J Neurophysiol. 2021. PMID 33978484
19
Lipovsek M, Marcovich I, Elgoyhen AB. The Hair Cell α9α10 Nicotinic Acetylcholine Receptor: Odd Cousin in an Old Family. Front Cell Neurosci. 2021. PMID 34867208
20
Wallace MT, Wilkinson LK, Stein BE. Representation and integration of multiple sensory inputs in primate superior colliculus. J Neurophysiol. 1996. PMID 8871234
21
Avillac M, et al.. Reference frames for representing visual and tactile locations in parietal cortex. Nat Neurosci. 2005. PMID 15951810
22
McKenna GJ, et al.. Neck muscle vibration alters visually perceived roll in normals. J Assoc Res Otolaryngol. 2004. PMID 14569429
23
Karnath HO, Christ K, Hartje W. Decrease of contralateral neglect by neck muscle vibration and spatial orientation of trunk midline. Brain. 1993. PMID 8461972
24
Beinert K, Keller M, Taube W. Neck muscle vibration can improve sensorimotor function in patients with neck pain. Spine J. 2015. PMID 25452010
25
Haavik-Taylor H, Murphy B. Cervical spine manipulation alters sensorimotor integration: a somatosensory evoked potential study. Clin Neurophysiol. 2007. PMID 17137836
26
Niazi IK, et al.. A randomized controlled trial comparing different sites of high-velocity low amplitude thrust on sensorimotor integration parameters. Sci Rep. 2024. PMID 38216596

Sources: primary literature, linked inline.

← All 61 lessons