The Frontal Lobe: Cortical Bands, Prefrontal Processes, and Release Signs

01Three frontal bands

The frontal lobe is built as three bands, and each band reaches movement by a different route

Primary motor cortex sits farthest back, Brodmann area 4, also called M I. Its giant pyramidal neurons project to anterior horn cells through the corticospinal tract and drive discrete voluntary movement. Anterior to it lie the premotor and supplementary motor cortices, both inside area 6. Prefrontal association cortex fills everything in front of them. Each band arrives at movement by its own route: area 4 straight down the cord, area 6 through the cue and the sequence, prefrontal cortex through what it will not permit.

The giant cells of area 4 are a measured class. Of 617 cortical neurons traced across 19 mammalian species from 7 orders, 181 met the criteria for gigantopyramidal neurons, the cells Betz described Jacobs 2018. They carry large somata and extensive basilar dendrites, and most dendritic measures exceed those of the superficial layer III and deep layer V pyramidal neurons quantified beside them. Gigantopyramidal size was greatest in feliforms, which the authors relate to muscle fiber and musculoskeletal specialization.

Area 4 is not the only road to the spinal cord

Six premotor areas project directly to the spinal cord alongside area 4. They sit in area 6 and in subfields of areas 23 and 24 in the cingulate sulcus. Together they account for more than 60 percent of the frontal cortex projecting to the cord in monkeys Dum 1991. Their combined corticospinal neuron count equals or exceeds that of the arm representation in M I. Each premotor area can therefore influence movement without passing through M I, which the authors say raises serious questions about calling M I the final common pathway.

The two frontal motor fields in area 6 fail in different ways

The lateral premotor region of the frontal lobe binds a sensory stimulus to a motor response. Patients with premotor lesions learned to pair 6 visual, tactile, or auditory stimuli with 6 previously rehearsed arm movements, and a comparison task paired the same stimuli with 6 spatial locations. They failed only when a movement had to be recalled from memory on the basis of a sensory cue Halsband 1990. The association involving spatial location survived intact. The deficit sits in the link between an arbitrary cue and an action.

The supplementary motor cortex, M II, sequences a motor plan already learned. Two patients with left medial lesions involving the SMA reproduced rhythms under auditory pacing and had the most severe difficulty producing any rhythm from memory, against 7 age-matched controls Halsband 1993. Lateral premotor lesions impaired rhythm reproduction as well, worst with alternating hands.

Prefrontal cortex is the largest heteromodal territory in the brain

Prefrontal cortex connects reciprocally with the parietal, temporal, and limbic cortices. Those connections fold perception, memory, and emotion into one output. Structurally it parses into the superior medial prefrontal cortex that contains the anterior cingulate, the lateral prefrontal cortex, the orbitofrontal cortex, and the frontal poles. Those four divisions hold separate control processes at separate addresses, which is why one frontal lesion produces apathy and another produces disinhibition. They cooperate closely enough that intention, sequence, and restraint feel like a single act of will.

02Findings

What the research shows

03Five prefrontal processes

Prefrontal control runs as five separable processes with separate addresses, and there is no central executive

Energization runs on the superior medial prefrontal cortices bilaterally. Task setting runs on the left lateral frontal cortex and monitoring on the right lateral prefrontal cortex. Behavioral and emotional regulation runs on the orbitofrontal cortex, and metacognition on the frontal poles Henri-Bhargava 2018. Those are the five prefrontal control processes that lesion work has managed to pull apart and localize.

Only task setting and monitoring count as executive functions, and every one of the five is prefrontal Henri-Bhargava 2018. Donald Stuss and colleagues reached the same three core relationships from lesion work: energization at superior medial cortex, task setting at left lateral cortex, and monitoring at right lateral cortex Stuss 2007. Those coarse localizations replicate across different tasks, cognitive modalities, and patient groups.

The sharper claim comes from the same body of work. Stuss described 5 anatomical and functional frontal relationships, adding the ventral-medial orbital region and the frontal pole to the three above Stuss 2011. What acts is a set of domain-general processes distributed across several frontal regions working in concert. Control is the outcome of several addresses cooperating, and naming it a faculty hides the addresses.

Energization sets how much output an intention produces

Energization is the initiation and sustaining of any nonreflex response. When it fails the result ranges from apathy and abulia to akinetic mutism, in which a patient lies awake, silent, and still. Phonemic word fluency measures energization through the count. The FAS test asks a patient to generate words beginning with F, A, and S.

Tests of phonemic and semantic fluency given to 53 patients with focal frontal-lobe lesions, 23 with unilateral temporal-lobe lesions, and 55 matched controls separated two components of the score Troyer 1998. Patients with left dorsolateral or superior medial frontal lesions switched between subcategories less often while producing cluster sizes of normal length. Phonemic-fluency switching was impaired only with frontal lesions, which is what licenses the FAS test as a frontal probe.

A larger fluency series splits the score further. Across 8 fluency tasks in 69 focal frontal patients, 43 posterior patients, and 150 controls, the superior medial region again carried energization and the left inferior frontal region carried selection Robinson 2021. The highest error rates came from lateral patients, not medial ones. Perseverative errors followed right frontal damage and phonemic rule-breaks followed left lateral damage. The count tracks energization, and the error type separates task setting from monitoring.

The two lateral frontal cortices work against each other

Inhibitory rTMS over the left lateral frontal cortex significantly worsened phonemic fluency against sham in 44 right-handed healthy adults, and the same inhibition over the right lateral frontal cortex significantly improved it Smirni 2017. The target was Brodmann area 47 on both sides. The authors attribute the right-side gain to release of interhemispheric inhibition. One input reached one target region and produced opposite results by side, which is what a single balance held between two cortices predicts.

04Working memory in area 46

Brodmann area 46 holds a location on line, and damage there produces a blind spot in memory rather than in vision

Working memory in the frontal lobe has a cellular address. Of 288 neurons recorded in and around the principal sulcus of monkey dorsolateral prefrontal cortex, 87 showed significant delay-period activity across delays of 1 to 6 seconds Funahashi 1989. Of those 87, 79 percent were directional. They responded only to cues from a certain range of visual field directions and weakly or not at all to others. Each such cell holds one place and stays quiet for the rest.

Lesions in the same region produce a matching failure. In 4 rhesus monkeys, dorsolateral prefrontal lesions rarely altered performance at the shortest delay of 1.5 seconds and made it progressively worse as the delay lengthened toward 8 seconds Funahashi 1993. Saccadic reaction times and velocities were unchanged, so seeing and moving were intact. The authors named the result a mnemonic scotoma.

The scotoma has a side. Each hemisphere handles the contralateral hemifield, which matters for the white matter, because the bundle carrying spatial information into area 46 arrives from the parietal lobe of the same side.

05The frontal white matter

The frontal lobe is defined by its bundles, and the human bundles do not match the monkey scheme

The superior longitudinal fasciculus links frontal cortex to parietal cortex in a bidirectional dialogue. In vivo diffusion tensor imaging separates four subdivisions in humans: SLF I, SLF II, SLF III, and the arcuate fascicle Makris 2005. SLF I lies in the white matter of the superior parietal and superior frontal lobes and extends to dorsal premotor and dorsolateral prefrontal regions, where it serves motor behavior under conditional rules. SLF II carries spatial perception from the angular gyrus to caudal lateral prefrontal regions, letting working memory in area 46 inform where attention falls. SLF III joins the supramarginal gyrus to ventral premotor and prefrontal cortex, knitting articulation and gesture together.

One limit belongs with those SLF trajectories. Diffusion tensor imaging defines the stem of a bundle, and the origins and terminations in that work were extrapolated from non-human primate material Makris 2005.

SLF I is the weak member of the scheme in humans

Fibers corresponding to SLF I appeared in only 6 of 20 human hemispheres, and in every one of those cases they belonged to the cingulum fiber system Wang 2016. Diffusion spectrum imaging in 10 adults, two group templates, and 5 autopsy dissections produced that result. The same work confirmed SLF II from the angular gyrus to the caudal middle frontal and dorsal precentral gyrus, and SLF III from the supramarginal gyrus to the ventral precentral gyrus and pars opercularis. The asymmetries it found are specific. Supramarginal connectivity with dorsal precentral and caudal middle frontal gyrus appeared only in the left hemisphere. SLF III connectivity with pars triangularis was constant only on the right.

Limbic bundles carry the frontal lobe's other half

The uncinate fasciculus connects the orbitofrontal cortex with the hippocampus and amygdala of the temporal lobe. Diffusion tensor imaging at 1.5 T with 41 gradient directions at b = 700 s/mm² measured the uncinate fasciculus in healthy right-handed adults. Fractional anisotropy ran left greater than right in the subinsular uncinate and inferior occipitofrontal fasciculi, with the same asymmetry in the arcuate Rodrigo 2007. The asymmetry tracks language specialization and was measured in the subinsular portion of the bundle.

The cingulum projects from the cingulate gyrus to the entorhinal cortex and receives thalamic afferents linked to pain processing. In 20 healthy controls and 25 people with amnestic mild cognitive impairment, cingulum microstructure predicted cognitive control, and the segments split by content Metzler-Baddeley 2012. Left anterior cingulum tracked control based on verbal or symbolic rules. Right anterior cingulum tracked spatial-rule errors. The anterior cingulate is one node in a distributed network.

Together these tracts form the frontoparietal network responsible for control instantiation, the system that holds a task in mind and steers behavior toward its completion. The far end of that network is the parietal lobe and attention.

06When frontal control fails

Frontal injury produces syndromes as varied as the regions damaged, and much of what appears is release

Orbitofrontal lesions yield behavioral disinhibition, emotional dysregulation, and altered social cognition. The findings include environmental dependency, loss of social decorum, inappropriate jocularity, and narrowed preoccupations. Superior medial lesions strip away energization instead, producing apathy, abulia, or in the severe case akinetic mutism. Disinhibition and apathy come out of one lobe, and which one appears is decided by a few centimeters of cortex.

Environmental dependency has a naming study. Two patients with focal unilateral frontal lobe lesions were observed across everyday settings Lhermitte 1986. Observation covered a doctor's office, a lecture room, a car, and a garden. It continued in an apartment where various activities were possible and in a gift shop. The patients imitated and used whatever each setting offered, as though implicit in the environment was an order to respond. The environment supplies the plan when the frontal lobe stops supplying it.

Phineas Gage was a disconnection

Phineas Gage, the railroad foreman who in 1848 survived a tamping iron through his prefrontal cortex, kept memory and speech and lost judgment. Measurement of his skull combined with modern neuroimaging placed the damage in both left and right prefrontal cortices Damasio 1994. His modern counterparts confirm the pattern, which produces a defect in rational decision making and the processing of emotion. His death is dated to 1861 and no autopsy was performed.

Applying the same lesion to a white matter atlas built from diffusion tractography of 129 healthy adults extended the disruption to connections projecting to areas distant from the lesion Thiebaut de Schotten 2015. The damaged tracts link areas engaged in emotion and decision making. Gage's personality change follows from disconnection of uncinate and cingulum traffic, which is why judgment failed while speech survived.

Impulsivity, perseveration, and a failing antisaccade are the release signs

A structured frontal exam finds impulsivity, rigidity, perseveration, and decreased affect. Loss of abstract thinking appears alongside hypomimia, slowed and impaired vertical saccades, and impaired antisaccades. The Luria hand sequence asks a patient to repeat a palm, fist, edge pattern, and it exposes difficulty in task setting. Go/No-Go and Stroop tasks probe inhibition. The exam is built to catch what has been let go, not what has been lost.

The antisaccade sign is precise about what has broken. Three patients with unilateral Brodmann area 46 lesions showed a bilateral increase in the percentage of misdirected reflexive saccades Pierrot-Deseilligny 2003. Visually guided saccades with gap and overlap were normal. So were the latency of correct antisaccades and smooth pursuit gain. The eye movement machinery is intact, and what the lesion breaks is the suppression of a saccade the patient has decided against. The saccade and pursuit systems then execute the unwanted movement faithfully.

Reemergent primitive reflexes make the same point at the arm. Grasp and glabellar tap signal release of frontal suppression over older subcortical programs. In 229 patients studied prospectively after brain damage, the elbow flexion response was significantly more marked in those with frontal lobe involvement than in those without Sudo 2002. It was stronger with bilateral than unilateral lesions and appeared more often with lower intelligence scores. The authors propose it is a proximal variant of the grasp reflex.

07The last cortex to mature

The frontal lobe finishes last, and its timetable is measured in synapses, gray matter, and decades

Synaptic density in human frontal cortex rises through infancy and peaks at age 1 to 2 years, about 50 percent above the adult mean. Direct counts in layer 3 of the human middle frontal gyrus ran across 21 brains from newborn to age 90. The adult mean for ages 16 to 72 was 11.05 × 10⁸ synapses per cubic millimeter, declining to 9.56 × 10⁸ in the aged Huttenlocher 1979. The surplus built by age 2 is pruned across childhood and adolescence.

Frontal timing runs out of step with sensory cortex. Maximum synaptic density in the middle frontal gyrus is not reached until after age 15 months, while auditory cortex peaks near 3 months Huttenlocher 1997. Net synapse elimination ends by age 12 in auditory cortex and extends into midadolescence in prefrontal cortex. Frontal development is heterochronous, and the offset is largest against the cortices whose output it integrates.

The volumetric sequence matches the cellular one

Higher-order association cortices matured only after the lower-order somatosensory and visual cortices whose functions they integrate Gogtay 2004. Thirteen healthy children scanned every 2 years for 8 to 10 years mapped gray matter between ages 4 and 21. Phylogenetically older areas matured before newer ones. Frontal cortex that integrates parietal and temporal output therefore waits on the cortex supplying it.

Frontal gray matter volume peaks at about 11.0 years in girls and 12.1 years in boys, and the dorsolateral prefrontal cortex does not reach adult dimensions until the early 20s Giedd 2004. White matter increases in a roughly linear pattern across all four lobes, with only minor differences in slope, while cortical gray follows an inverted U with much greater regional variation. The frontal lobe's white matter keeps gaining volume across adolescence while its gray matter is already past peak.

Metacognition sits in the frontal poles, the most anterior cortex in the brain, and behavioral regulation sits in the orbitofrontal cortex. The capacity to weigh long-term consequence against immediate reward arrives gradually because the dorsolateral cortex holding it is still under construction into the early 20s. A 15-year-old runs the same five prefrontal processes as an adult on a substrate that is still being cut down.

08Reaching the frontal lobe

Input from the spine changes how the prefrontal cortex processes somatosensory information

In 19 adults with subclinical pain recorded on 62-channel EEG, spinal manipulation of dysfunctional segments dropped the somatosensory evoked N30 amplitude by 16.9 percent, with no change after the control intervention Lelic 2016. Brain source modeling resolved a four-source model, and only the prefrontal source moved, falling by 20.2 percent. The other three sources held still.

The larger trial repeats the fall. Ninety-six adults with recurrent mild neck pain were randomly allocated under double blinding to a single thrust at a clinician-chosen relevant segment or at a predetermined non-relevant one. The relevant arm dropped the N30 by 16.76 percent at p equal to 0.005, and the non-relevant arm did not move, at p equal to 0.757 Niazi 2024. Both samples were neurologically intact and both fell, which is the established direction after a spinal input. Which segment the clinician chose is the subject of the research page on tone.

A cortex carrying a chronic lesion answers in the opposite direction

In a randomized crossover trial, 17 men with chronic stroke, mean age 53 years, received the same class of input. The N30 rose by 39 percent against the control intervention, at p below 0.01 Navid 2020. The N20 peak, the resting-state power spectra, the delta-alpha ratio and the brain-symmetry index were all unchanged. That is the one published reversal of the sign, and it is the one sample whose cortex carried structural damage.

Both directions were already inside the falling groups. The standard deviation on the N30 change after spinal manipulation was 31.3 percent around a mean of 16.9 percent in the 19-adult study Lelic 2016. In the 96-adult trial it was 28.32 percent around a mean of 16.76 percent Niazi 2024. Each spread runs close to twice its own mean, so frontal lobes inside a group that fell on average did not all fall. The input was held constant across those samples and the cortex meeting it was not.

That is the prediction the Unified Model of Tone makes for this cortex. Correspondence between the input and the suppression the frontal lobe is already holding decides the size and the sign of the change, which is why a heavier contact buys no larger prefrontal shift. The model's own test is stricter than any of these trials ran. A cortex holding too much suppression and a cortex holding too little should move toward the same middle on the same information. The spread of prefrontal source activity across a sample should therefore narrow. A standard deviation of 31.3 percent around a mean of 16.9 percent is what heterogeneous starting states produce. The narrowing itself is measured from the before and after spread, and that is the comparison none of these studies reports.

The contact is made at a joint and the change appears in the frontal lobe

Chiropractic care acts on spinal segments, and in both of these recordings the frontal lobe is where that contact was measured, at the prefrontal source of a four-source model. Health in this lobe is a restraint that can be held and then lifted on time. Tone drifted outside that range shows up as an intention that drives too hard or not hard enough, and as behavior released before the moment fits.

09Tone

How this system expresses tone

The frontal lobe expresses all of tone, and its signature is legible in what never happens. What this lobe holds down, how hard it drives an intention, and how its two lateral cortices pull against each other are the readings it gives.

Each remaining foundation shows up in this lobe as something held or something let go. Set point: prefrontal cortex defends a working range, which is why area 46 lesions show up as delay-dependent failure instead of an all-or-nothing loss. Prediction: area 46 lesions cut the percentage of anticipatory saccades in a predictive task bilaterally, so what the lesion takes is the plan made before the target appears. Load: holding a location available costs continuous firing, which is what 87 principal sulcus neurons were doing across delays of 1 to 6 seconds. Oscillation: energization rises to launch a nonreflex response and falls when the task ends, so frontal drive arrives in bouts. Time course: pruning that runs into midadolescence sits under a dorsolateral cortex that reaches adult dimensions only in the early 20s, and both sit under a lesion effect measured in a single session. Input quality: the prefrontal source answered a spinal joint, falling 20.2 percent after manipulation of dysfunctional segments.

10Across the library

How this page relates to the rest of the library

11Frequently asked

Questions about this topic

12The sources

References

Sources: primary literature, linked inline.