The Motor Unit and the Final Common Path

A motor unit is defined as a single alpha motor neuron together with all the extrafusal skeletal muscle fibers it innervates. This is the smallest unit of controllable force: because a motor neuron either fires an action potential or it does not, and because that action potential triggers the neuromuscular junction to release acetylcholine onto every one of "its" fibers simultaneously, all fibers within a motor unit contract together, in an all-or-none fashion. You cannot activate half a motor unit. Fine control of force therefore does not come from grading the contraction of individual fibers but from controlling how many motor units are active and how fast each one is firing. The number of muscle fibers served by one motor neuron is its innervation ratio, and this ratio varies by more than a thousandfold across the body. Muscles built for precision — the extraocular muscles that aim the eye, or the small muscles of the larynx and fingers — have innervation ratios as low as 1:3 to 1:10 fibers per neuron. Muscles built for gross force — the medial gastrocnemius or soleus of the leg — have innervation ratios in the range of 1:1000 to 1:2000. A muscle is thus not one thing electrically; it is a mosaic of hundreds of independent motor units, each a discrete cellular franchise with its own axon, terminal arbor, and cluster of fibers scattered (not clumped) throughout the muscle's cross-section. Sherrington's phrase "final common path" captures a key architectural fact: the alpha motor neuron receives thousands of synaptic inputs — direct monosynaptic connections from corticospinal and rubrospinal neurons, polysynaptic relays through spinal interneurons, proprioceptive afferents from muscle spindles and Golgi tendon organs, and inputs from central pattern generators — yet it has only one axon and one output. Every plan for movement, however it was computed upstream, must be translated into the language of this single cell's firing rate. This convergence makes the motor neuron simultaneously the simplest and the most consequential node in the The Motor Hierarchy: From Cortex to Muscle.

Step 1: Anatomical convergence. Each alpha motor neuron cell body sits in the ventral horn of the spinal cord (or in cranial motor nuclei for head and neck muscles), organized into longitudinal columns — motor neuron pools — each pool dedicated to one muscle. The pool for a given muscle can contain anywhere from a few dozen to several hundred motor neurons, spanning several spinal segments.

Step 2: One neuron, many fibers, one twitch type. Each motor neuron's axon leaves the cord, travels in a peripheral nerve, and arborizes near its target muscle, sending a terminal branch to every fiber in its motor unit via the Neuromuscular junction. Critically, all fibers within one motor unit share the same physiological properties — a motor neuron enforces a uniform contractile phenotype on its fibers through trophic and activity-dependent signaling. Three broad types exist:

Step 3: The size principle governs recruitment order. When the motor neuron pool for a muscle receives an excitatory drive (from descending tracts or reflex afferents), that drive is delivered roughly equally, as a graded synaptic current, to all motor neurons in the pool. But motor neurons do not all reach firing threshold together. Small motor neurons (type S) have a small cell body and correspondingly high input resistance (by Ohm's law, V = IR, a given synaptic current I produces a larger depolarization V when R is large). This means small motor neurons reach threshold first, at the lowest levels of synaptic drive. As drive increases, medium (FR) motor neurons are recruited, and only at high drive are the large (FF) motor neurons — with their low input resistance — finally brought to threshold. This produces an invariant order: S before FR before FF, regardless of the behavioral task, almost always. This ordering is elegant because it is automatic: it emerges purely from passive membrane biophysics (soma size, input resistance, axon diameter, and afterhyperpolarization duration), not from any separate "decision" the nervous system makes about which units to call up. The descending and reflex systems only need to control one variable — the overall level of excitatory drive to the pool — and orderly recruitment falls out for free.

Step 4: Two independent knobs for force — recruitment and rate coding. Total muscle force is graded by: 1. Recruitment — adding more motor units, starting small and weak, ending large and powerful. This dominates control at low-to-moderate forces (roughly the first 50–90% of a muscle's force range, depending on the muscle). 2. Rate coding — increasing the firing frequency of already-recruited motor units. A single action potential produces a brief twitch; if the next action potential arrives before the muscle fully relaxes, twitches summate mechanically (temporal summation), and at high enough firing rates the individual twitches fuse into smooth, sustained tetanus. Rate coding becomes progressively more important as recruitment approaches its ceiling, especially in muscles with few, large motor units (like hand muscles) where almost all units are recruited early and further gradation depends heavily on firing rate.

Step 5: Why this design makes ecological sense. Small, slow, fatigue-resistant motor units are recruited constantly, for posture and low-force, sustained tasks — they need to resist fatigue because they are almost always "on." Large, fast, powerful, fatigable motor units are recruited only rarely, for brief, forceful bursts (jumping, a maximal grip) — they can afford to fatigue quickly because they are used sparingly. The size principle thus matches the recruitment order to the metabolic economy of the muscle: cheap, durable units do most of the everyday work, and expensive, high-output units are held in reserve for emergencies.

The size principle was established through a research program led by Elwood Henneman and colleagues at Harvard Medical School beginning in the 1960s. Their central technique was intracellular and extracellular recording from spinal motor neurons and their axons in decerebrate and anesthetized cats, correlated with the size of the recorded cells.

The core experiment. Henneman's group recorded from motor axons in the ventral roots or from motor neuron somas in the spinal cord innervating hindlimb muscles (notably the triceps surae group — soleus and gastrocnemius) while gradually increasing a natural excitatory input, such as stretch of the muscle's own spindle afferents (a monosynaptic reflex) or graded electrical stimulation of afferent nerves. They measured two things for each recorded motor neuron: its axonal conduction velocity (a proxy for axon diameter, and hence cell size, since larger cells have larger axons) and the stimulus intensity at which it began firing (its recruitment threshold).

What was observed. Across hundreds of motor neurons and many preparations, the relationship was remarkably orderly and monotonic: motor neurons with the smallest cell bodies and slowest-conducting axons had the lowest recruitment thresholds, firing first as stimulus strength increased from zero. Motor neurons with progressively larger somas and faster-conducting axons had progressively higher thresholds and were recruited later, in an almost perfectly rank-ordered sequence. This held true whether the excitatory drive came from stretch reflexes, from stimulating descending pathways, or from voluntary contraction in later human studies — the order of recruitment for a given pool was essentially the same regardless of how the pool was activated, a finding Henneman used to argue that recruitment order is a fixed property of the motor neuron pool's biophysics, not a centrally programmed strategy that changes with task.

Corroborating human evidence. In humans, intramuscular fine-wire and needle electromyography (EMG) studies from the 1960s onward (including seminal work by Milner-Brown, Stein, and Yemm) recorded single motor unit action potentials during graded voluntary contractions. As subjects increased force smoothly from a gentle contraction to a maximal one, investigators observed new, larger-amplitude motor unit potentials being recruited in an orderly sequence, coincident with increases in the firing rate of already-active units — direct human confirmation that both recruitment (in size order) and rate coding jointly grade force, exactly as the cat electrophysiology predicted.

What it proved, precisely. These experiments established three things: (1) that motor units differ systematically in a rank-orderable "size," correlated across soma diameter, axon conduction velocity, twitch force, and fatigue resistance; (2) that recruitment threshold correlates tightly and inversely with that size, consistent with the input-resistance argument from cable theory; and (3) that this order is largely invariant across different modes of activation, implying the ordering is built into the passive electrical properties of the motor neuron rather than actively selected trial by trial by upstream circuits. Later work (Zajac, Faulkner, Burke) refined the picture by classifying motor units into the S/FR/FF scheme based on twitch and fatigue properties, showing these histochemical and physiological types map onto the same size-ordered recruitment sequence. It is worth noting the size principle is not absolute law: a small number of documented exceptions exist (for example, some fast "ballistic" or specialized tasks and certain reflexes can partially reorder recruitment), and the biophysical account has been refined by contributions from persistent inward currents and neuromodulation. But as a first-order description of vertebrate skeletal motor control, it has held up for over half a century of scrutiny.

Example 1 — Innervation ratio and force resolution. Suppose the extensor digitorum brevis (a small foot muscle) has 200 motor units and 4,000 muscle fibers total, while the medial gastrocnemius has 300 motor units and 900,000 fibers total. Average innervation ratio:

• EDB: 4,000 / 200 = 20 fibers per motor neuron
• Gastrocnemius: 900,000 / 300 = 3,000 fibers per motor neuron

If we assume (simplifying) that each fiber contributes equally to force, then the smallest controllable force increment — the force added by recruiting just one more motor unit — is roughly 1/200 of maximum in the EDB but only 1/300 of maximum in the gastrocnemius, in the naive case where all units are equal-sized. But real motor unit sizes are not uniform — the size principle predicts they increase progressively, so the smallest (first-recruited) units contribute a much smaller fraction of total force than this crude average suggests, while the largest (last-recruited) units contribute disproportionately more. This is precisely why the small units are recruited first: recruiting them early gives fine force resolution near the bottom of the range, exactly where fine control (e.g., holding a pencil lightly) matters most, whereas the coarse jumps produced by the huge FF units late in recruitment are acceptable because at high forces (e.g., a maximal jump), fine gradation is behaviorally less important than raw power.

Example 2 — Rate coding and tetanic fusion. A motor unit's twitch has a time-to-peak of 40 ms and a half-relaxation time of 50 ms (so the twitch has largely decayed by about 100 ms after the stimulus). Consider firing rates of 5 Hz, 15 Hz, and 40 Hz.

• At 5 Hz, the interspike interval is 1000/5 = 200 ms — much longer than the twitch duration, so each twitch is discrete and force output oscillates between individual twitches (unfused, low mean force).
• At 15 Hz, the interspike interval is 1000/15 ≈ 67 ms — shorter than full twitch decay, so twitches begin to summate; force rises and shows ripples (partially fused tetanus), with mean force substantially higher than a single twitch.
• At 40 Hz, the interspike interval is 1000/40 = 25 ms — much shorter than the twitch's rise time, so successive twitches summate almost completely before the previous one decays, producing smooth, fused tetanic force close to the motor unit's maximal tetanic tension (often 3–5× the peak twitch force for a single unit).

This illustrates why increasing firing rate (rate coding) is such an effective way to grade force even after a motor unit is already recruited: the same motor unit, driven from 5 Hz to 40 Hz, can change its mean force output several-fold, entirely without adding new units. In muscles where only a handful of large motor units exist (such as some small hand muscles), rate coding — not recruitment — necessarily does most of the fine-force work once the available units are largely recruited.

Understanding motor units and the size principle underlies much of modern clinical neurophysiology and rehabilitation science. Electromyography (EMG), used to diagnose diseases like amyotrophic lateral sclerosis (ALS), peripheral neuropathy, and myopathy, interprets abnormal motor unit potentials — increased amplitude and duration signal collateral reinnervation after motor neuron loss, while small, short-duration, polyphasic potentials suggest primary muscle disease — directly on the basis of motor unit anatomy described here. In ALS, progressive loss of large, fast-conducting motor neurons (paradoxically often the largest FF units are among the most vulnerable) helps explain the characteristic early loss of maximal force and fine control even while some function is preserved by compensatory reinnervation from surviving small units. In prosthetics and neural interfaces, engineers exploit the size principle and rate coding to decode intended force and movement from residual EMG signals in amputees, since the orderly relationship between drive, recruitment, and firing rate gives a predictable, decodable structure to muscle electrical activity. In exercise physiology and strength training, the size principle explains why heavy, near-maximal resistance training preferentially engages and adapts the large, fast, fatigable motor units (driving hypertrophy and power gains), while light, high-repetition endurance training predominantly works the small, fatigue-resistant units — informing evidence-based training prescriptions for athletes and rehabilitation patients alike. In spinal cord injury and stroke rehabilitation, clinicians track changes in recruitment order and motor unit firing patterns (using surface EMG decomposition) as a biomarker of motor recovery and reorganization, since disruptions to descending drive can alter the orderliness of recruitment even when the peripheral motor units themselves remain healthy. Finally, in botulinum toxin and neuromuscular pharmacology, dosing decisions depend on knowledge of typical innervation ratios in target muscles (small ratios in facial and extraocular muscles requiring far smaller, more localized doses than large limb muscles).

Ask the tutor: How does the size principle interact with descending pathways during voluntary versus reflexive movements — does the cortex ever bypass orderly recruitment?