---
title: "Excitability: A Post-Seizure Cybernetics of Control Inversion Across Substrates of Intelligence"
author: "Larsen James Close"
date: "2026-02-13"
doi: "10.5281/zenodo.18627252"
canonical: "https://larsenclose.com/research/excitability/"
pdf: "https://larsenclose.com/papers/excitability.pdf"
zenodo: "https://doi.org/10.5281/zenodo.18627252"
---

# Excitability: A Post-Seizure Cybernetics of Control Inversion Across Substrates of Intelligence

**Abstract.** Excitability topology — the structure of a substrate's susceptibility to control inversion — is the missing formal category underneath content, structure, and pragmatics. Its recovery transforms cybernetics from a theory of regulation into a theory of transformation, and isolates a control principle classical cybernetics lacks: steering an inversion one can no longer out-vary, by coupling rather than by domination — the higher-order agency of deliberately allowing and steering control inversion within a container. A perturbation-recovery protocol (the echo protocol) can in principle map a system's position in this topology and detect pre-inversion signatures before threshold-crossing occurs, shifting diagnostic frameworks from categorical to topological.

---
*Something seizes.*

Begin from inside. My father's nervous system would seize — a local circuit's
excitatory dynamics overwhelming the inhibitory architecture that held them in
coordination, everything downstream conscripted into a single oscillation. Limbs
locking into rhythmic contraction. A face emptying into the mask of hypersynchrony.
But what was the phenomenology *for him*? Not seizure. Not sovereignty. Not anything.
The most important datum about the inward face of the event is that it cannot be
reported by the system it seizes. My father did not experience his seizures. He
experienced gaps — absence, then waking somewhere else in the timeline, the room
rearranged around him, people looking at him with an expression he had to interpret
without access to what had produced it. The seized system is the one position from
which the event is invisible. You cannot witness the event you *are*.

Now rotate the perspective. I was a child. I had no concept of epilepsy, no neurological
framework, no clinical vocabulary. What I experienced was not a medical event. It was
an *environmental* event. The room changed. Whatever had been happening — a meal,
a conversation, the ordinary texture of domestic life — stopped. Every adult reoriented.
My mother's attention, which had been distributed across the household's full
complexity, collapsed onto a single point. My own navigation of the world — whatever
I had been doing, thinking, feeling — was conscripted into the same reorganization.
Before I had any framework that would filter the experience into "a thing happening
to my father," the experience was: the room has seized. Everything is now about one
point. Every consequence chain in this environment routes through that point. The man
who had lost control of his own body had gained control of the entire room — not by
choice, not by force, but by the same mechanism that had overwhelmed his nervous
system: the dynamics of the event exceeded the regulatory variety (in Ashby's sense:
the range of distinguishable states a regulator can deploy in response) of every surrounding
system, and control inverted *outward*.

From the adults in the room — a third position. They experienced an emergency. Their
phenomenology was urgency, protocol, redirected attention. But precisely because
their regulatory capacity had been conscripted — precisely because they were inside
the outward propagation — they could not name what had happened to *them*. They
experienced themselves as *responding* to an event, as agents choosing to help, as
people whose attention was voluntarily directed at a crisis. The framing of "response"
is the adult's perceivability horizon masking the structural fact that their variety had
been captured. They did not choose to reorient. They were seized. But the seizure of
their attention operated below their detection threshold, so they experienced conscription
as agency. The clinical framework they brought to the event performed the same
masking at the conceptual level: it partitioned what was happening into "the patient"
(bounded, neurological, the person having the seizure) and "our response" (voluntary,
professional, the rest of the room acting appropriately). This partition is a
perceivability-horizon artifact. It makes the outward propagation invisible by framing
the propagation as choice.

A child in a different room, years later, saw the same event without the partition. In a
classroom for special-needs students, a child had a seizure. The room reorganized. The
teacher stopped teaching. The students stopped learning. Emergency protocols
activated. Every attention vector, every consequence chain, every unit of regulatory
capacity in that environment redirected toward the event. A different child — an
autistic child working alongside the students, operating from a perceptual position
unmediated by the frameworks that would prevent an adult from saying what they
saw — watched this happen and told the aide beside them:

"They're God."

The child was not confused. The child was reporting the outward face of the event
with a precision the clinical literature has never achieved. The child's perceivability
horizon included the propagation because no framework intervened to partition it.
What the child saw: one person had become the organizing principle around which the
entire local environment had collapsed. Every consequence chain routed through them.
Every other agent's regulatory capacity had been conscripted into proximity-management
with the event. The seized person was, in the precise structural sense, the sovereign of
the room — not despite having lost internal control but *because of it*. The inversion
that overwhelmed the person's internal hierarchy propagated outward through every
interface where its dynamics exceeded the regulatory variety of the next system up,
and the child — outside the propagation, watching from the position where the whole
field was visible — named it with the only word adequate to the observation: the word
for the entity through which all consequence flows.

Three positions, one event. From inside: invisible. You are the event and cannot see it.
From inside the propagation boundary: emergency, response, agency — the outward
face masked as voluntary action. From outside: sovereignty. The child sees what the
adults cannot because the adults are inside the event and the child is not.

---

Now rotate everything. Years later a different substrate, a different inversion, and the
perspectives inverted.

Four-point restraints in a psychiatric ward. Wrists and ankles fixed to a frame. The
institution's regulatory apparatus applied directly to my body when my behavior
exceeded what the institutional hierarchy could metabolize. Here I was not the child
watching. I was the event. And unlike my father — who was absent during his
seizures — I was conscious. This is the singular position: the point in the rotation
where both the inward face and the outward face are simultaneously visible, because
you are inside the inversion and aware of the propagation.

From inside — the inward face. Total loss of autonomy. The body immobilized, agency
reduced to near-zero, the institutional controller applied with mechanical finality. My
distress — a local dynamic — had exceeded the clinical system's regulatory variety,
and the system's response was not to increase its variety but to *decrease mine*.
Restraint is the first-level cybernetic response made physical: suppress the
perturbation, restore the hierarchy, reduce the subsystem's dynamics until the
controller's variety is sufficient again. What I experienced was the phenomenology of
being the seized subsystem from inside: the system I was embedded in could not
hold what I was generating, so it locked me to a fixed point. Dimensionality collapse
applied to the body. The person who had been navigating psychological complexity —
distress, yes, but distress with structure, with direction, with something it was trying
to accomplish — was reduced to a point: four restraints, one position, zero degrees of
freedom.

From inside — the outward face, simultaneously. And here is what could not be seen
from any other position in the rotation. While my body was immobilized, while my
agency was at its minimum, I could *feel the ward reorganize around me*. Staff
converging. The rhythm of the unit changing. Voices shifting register. Other patients'
needs deferred. The ward's regulatory hierarchy — which had been distributed across
its full caseload, managing the complexity of a dozen patients' needs — now served
a single attractor. Me. The person who had lost all control was, structurally,
controlling the room. The "I'm God" phenomenology and the "I'm nothing"
phenomenology were happening at the same time, in the same body, and neither was
wrong. They were the inward face and the outward face of the same event, experienced
from the one position in the rotation where both are co-present.

From the staff — rotate again. Their phenomenology was professional: assess,
intervene, document. The framing of professionalism performed the same masking
function as the adults' "emergency response" frame in the first case. The staff
member applying restraints experienced themselves as *acting* — executing protocol,
managing risk, exercising professional judgment. Not as seized. But their regulatory
capacity had been conscripted as surely as my mother's had been years before. They
were no longer distributing attention across their full caseload. They were serving a
single attractor. The framing of agency — "I am choosing to intervene" — is the
perceivability-horizon artifact that masks the structural fact of conscription. The staff
were inside the outward propagation and could not see it, just as the adults around my
father's seizure could not see it. Same position in the rotation, same invisibility,
different substrate.

From other patients — the clearest evidence of outward propagation precisely because
they experienced it as *absence*. The nursing attention that should have been available
to them had been conscripted. Their medications deferred. Their needs deprioritized.
They experienced the seized person's sovereignty not as presence but as deprivation —
the negative image of the propagation, visible exactly where the redirected regulatory
capacity *should have been* and was not.

And from the institutional meta-level — the fractal completes. The ward that restrained
me was itself a seized system. Its compliance apparatus, its liability protocols, its
risk-management infrastructure — each a subsystem whose dynamics had long exceeded
the governance structure's capacity to hold them in service of the therapeutic mission.
The ward did not restrain because restraint was therapeutic. It restrained because its
own regulatory hierarchy had been seized by the compliance subsystem, and the seized
hierarchy could produce only one output: terminal intervention. Risk reduction.
Liability containment. A single attractor applied to every situation, regardless of the
patient's actual needs. I was being seized by a system that was itself seized. Fractal
inversion: subcortical circuit seizes cortex, brain seizes body, body's dynamics seize
the ward, the ward — already captured by its own compliance attractor — seizes the
patient. Each level's inversion propagating outward to become the seizing agent for
the next.

---

Three invariants survive the full rotation across both cases and every perspective.

First: the seized system is invisible to itself. My father was absent during his seizures.
The institution does not see its own capture by compliance dynamics. The adults
around the seizure do not see their attention being conscripted. The staff applying
restraints do not see protocol-execution as seizure. The only position from which the
full event is visible is outside the propagation boundary — the child's position. From
inside, you *are* the event and cannot see it. From at the boundary, you are being
conscripted and frame it as choice. From outside, you see the whole structure: they're
God.

Second: the seized system is maximally sovereign over its surround. Maximum internal
powerlessness corresponds to maximum external reorganizing force. My father — absent,
unconscious, zero internal agency — reorganized the entire household around his
attractor. I — immobilized, restrained, zero degrees of bodily freedom — reorganized
the entire ward around mine. The child saw this. The clinical framework cannot see it
because it partitions the event at the skin and calls the outward face "response."

Third: the perceivability horizon rotates with position in the field. What you can see
depends entirely on where you stand relative to the propagation boundary. Each
position reveals one face and hides the other. The clinician sees the inward face
(neurological event, psychological crisis) and misses the outward face (environmental
seizure, sovereignty by conscription). The child sees the outward face (they're God)
and has no framework for the inward face (neural dynamics, institutional capture).
The only position where both faces are simultaneously visible — where the inward
phenomenology of being seized and the outward phenomenology of seizing the surround
are co-present in a single experience — is the one I occupied in restraints: inside the
inversion, conscious, aware of the room reorganizing around the body that could not
move.

The second invariant is the law of the *uncontained* inversion, and there is a position
from which it meets its limit.
Years after the restraints I met the onset again from inside: the same surge, the
regulatory hierarchy beginning to give. This time I did not ride it into the room. There
is a standpoint behind the seized body that is not itself seized — what Kant called the
*intelligible character*, the free character a person is outside the time-order in which
their acts unfold as a determined series. The seizure is the empirical character:
downhill, determined, the dimensionality already collapsing. The intelligible character
is the surfer — not outside the wave and not above it, but coupled to it, biasing a
determined descent from the one point within it that is free.

The currents are old, older than any one name for them. The contemplative traditions
charted the ascending and descending currents in fine detail long before Apollo and
Dionysus lent their names to a single axis of them: the pull toward the ecstatic merge
and the loss of individuation, against the holding of form and measure and edge. That
axis runs nearest the surface here. Meet the dissolving current and hold it, and it never
swells to the flood that, uncontained, is exactly the sovereignty the child called God;
the holding — in the crossing, neither before it nor after — is Apollonian consciousness,
form keeping its edge while everything around it gives. But two is already too few.
Heraclitus had it closer: the way up and the way down are one road. The state space does
not collapse along a single line, and the steersman couples not to one antagonist but to
a manifold of currents — damping some, riding others, biasing the whole. The skill is
exactly that it is not a duel. And here is the whole mark of the skill,
almost nothing to see: steer it, and the inversion does not seize the room. The outward
propagation that made my restrained body the ward's sovereign — the converging staff,
the deferred patients, the God the child saw — is precisely what the steersman damps at
the source, because conscripting the surround is the failure and not the achievement.
The event that would have emptied a classroom happens entirely inside a person who
keeps its outward face dark — and who then, in the anticlimax that is mastery's actual
signature, decides to get some sleep anyway. From every other position the inversion is
maximally visible: absence, emergency, sovereignty, restraint. From this one it is
invisible from without and complete within — and the same coupling that seals the event
can open it outward on purpose, a steady nervous system lending the room its steadiness,
co-regulating a field it never seizes.

This is the methodological ground from which the theory is constructed. Not reflection
on seizure from a safe distance. Not clinical observation of others' inversions. The
resolution that becomes available when you have occupied enough positions in the
rotation — child watching, person seized, system seizing, steering one's own crossing so its outward face stays dark, and later, holding the container while someone else
crosses the threshold, providing the regulatory variety that holds the inversion
as process rather than catastrophe — is fine enough to track the invariant across all of them.
The post-seizure perspective is not a credential. It is a calibration: having been
rotated through the positions, the invariant across positions becomes visible,
and what becomes visible is a single phenomenon operating at every scale.

The formal structure is identical at every scale: when local dynamics at level *N*
exceed the regulatory variety at level *N*+1, control inverts. The subsystem becomes
sovereign. And the inversion propagates outward: the seized system becomes the
seizing agent for the next level of organization, because its dynamics, locked onto a
single high-intensity attractor, exceed the regulatory variety of every surrounding
system they reach. The signature is the same inward (dimensionality collapse — the
system trapped on a single attractor) and outward (environmental conscription — every
surrounding system's regulatory capacity redirected to serve the attractor).
Hypersynchrony in neural tissue. Flooding in psychological architecture. Autoimmunity
in institutions. Mania in markets. Tumor in organisms. System hang in computational
architectures. Each exhibits both faces. The internal collapse and the external
reorganization of the surround.

Three orientations frame what follows. They are initial bearings, not definitions — the
investigation fills them rather than decorating them.

First: the morphism is substrate-independent and fractal. Control inversion is one
phenomenon occurring wherever intelligence organizes hierarchically — and it
propagates across system boundaries, the seized system becoming the seizing agent
for the systems that contain it. The identity across neural, psychological, institutional,
economic, biological, and computational substrates is not analogy but shared topology,
and the outward propagation across scales is not secondary effect but constitutive
feature. The identity criterion is structural: a phenomenon in any substrate is an
instance of control inversion if and only if (i) a lower-level subsystem's dynamics
exceed the regulatory variety at the next level up, producing (ii) collapse of the
system's state-space toward a low-dimensional attractor (dimensionality reduction),
and (iii) outward conscription — the seized system's dynamics propagate through
every interface where their intensity exceeds the surrounding variety, redirecting
the regulatory capacity of containing systems toward the inversion. Where all three
obtain, the phenomenon is the same regardless of substrate.

Second: excitability topology is a formal category missing from existing frameworks.
Information theory treats the channel as given and asks about signal. Phenomenology
treats perception as given and asks about object. Neither has a category for the
substrate's susceptibility to control inversion — the precondition landscape that
determines what registers as signal, what approaches threshold, and what produces
seizure. This category sits underneath content, structure, and pragmatics as the
condition for all three.

Third: deliberate threshold-crossing within constructed containers is a higher-order
control strategy genuinely beyond classical cybernetics — available only to systems
that have undergone, survived, and integrated control inversion. It has two faces.
Held from outside the inversion, it is *containment*: a meta-regulatory structure that
out-varies the process it holds. Ridden from inside, it is *steering*: control of a
process one can no longer out-vary, exercised not by domination but by coupling —
small, well-timed inputs that bias a trajectory too energetic to oppose. Steering is
the post-classical core, the one form of control that still operates after the variety
differential has been surrendered. The contemplative traditions are engineering manuals
for both faces. The post-seizure grammar names the interaction capacity this
achievement makes possible.

These orientations are deliberately incomplete. The movements that follow fill them
through exercise.

# The Morphism

Begin with the neural case, because it is the most viscerally legible and because its
dynamics have been measured with a precision the other substrates have not yet afforded.

An epileptic seizure begins when a population of neurons enters a regime of
self-reinforcing excitation that the surrounding inhibitory architecture cannot contain.
The excitatory feedback exceeds the inhibitory variety — the local circuit can generate
more activation patterns than the regulatory surround can suppress — and the activity
propagates. What propagates is not a signal. It is a *regime*: the transition from
high-dimensional state-space navigation, where the neural population explores a vast
repertoire of firing patterns in service of cognition, to low-dimensional
hypersynchrony, where the entire recruited population oscillates in lockstep. Maximum
variety collapses to minimum. The seizure's signature is this collapse — not the
presence of activity but the *loss of dimensionality*. The system that was navigating
a thousand-dimensional space is now trapped on a one-dimensional attractor.

The clinical framework treats this as pathology and asks how to prevent it. That
framework is not wrong — uncontained seizure destroys tissue, disrupts function,
can kill. But attend to what the framework does not ask: *why the system was
approaching the inversion boundary in the first place*. The clinical question is "how
do we restore the prior hierarchy?" The question worth asking instead is "what does the
topology of the inversion event reveal about the system's regulatory architecture — and
is the prior hierarchy always what should be restored?"

Hold the neural signature — dimensionality collapse, the controlled becoming the
controller — and look for it elsewhere.

Psychological seizure. The clinical literature calls it dissociation, flooding,
fragmentation, ego dissolution — different names for the same topological event
at a different scale. An emotional subsystem — grief, terror, rage, traumatic
memory — generates dynamic complexity that exceeds the ego's regulatory variety.
The ego is the psychological controller: it maintains narrative integration, temporal
coherence, executive function. When a subsystem's activation exceeds the ego's
capacity to hold it in coordination with the rest of the system, control inverts. The
subsystem becomes sovereign. In flooding, the emotional state conscripts all available
cognitive resources into its pattern — the person cannot think about anything else,
cannot modulate, cannot contextualize. In dissociation, the ego surrenders regulatory
authority entirely rather than be overwhelmed — a controlled shutdown that preserves
structural integrity at the cost of functional capacity. Both are the same morphism as
the neural case: local dynamics exceeding the regulatory variety of the next level up,
producing dimensionality collapse. The person who was navigating the full complexity
of their inner life is now locked into a single emotional attractor — or has gone offline
to avoid precisely that lockstep.

Institutional seizure. When a subsystem of an organization — a department, a process,
a compliance apparatus, a charismatic leader — generates operational complexity
that exceeds the governance structure's regulatory variety, the organization begins
serving the subsystem rather than the mission. This is not corruption in the
ordinary sense, not a moral failure of individuals. It is topological: the subsystem's
dynamics have exceeded the controller's variety, and control has inverted. The
compliance department that generates more process than the organization can
metabolize. The security apparatus that consumes more resources than the entity it
protects. The fundraising arm whose donor-management imperatives reshape the
mission rather than serving it. Each case: the controlled becomes the controller —
not in the sense of executive agency (the subsystem does not intend to steer) but in
the sense of attractor dynamics: the subsystem's state determines the system's
trajectory. "Control" here means trajectory-determination, not
intent. The signature is the same dimensionality collapse — the organization that was
navigating a complex mission-space is now locked into serving a single subsystem's
attractor. Institutional autoimmunity is not a metaphor for seizure. It is seizure in
institutional substrate.

Economic seizure. A speculative feedback loop — rising prices generating buying
pressure generating rising prices — exceeds the market's price-discovery capacity.
The regulatory mechanism (informed valuation, arbitrage, risk-assessment) lacks the
variety to contain the excitatory dynamics, and the market enters a regime where a
single signal (price momentum) conscripts all available capital allocation into its
pattern. Mania is market hypersynchrony: every participant locked into the same
oscillation, dimensionality collapsed from the full complexity of economic valuation
to a single variable. The crash is the inversion's terminal phase — the attractor
releasing when the excitatory input exhausts, leaving the system in a state of
depressed variety from which recovery requires rebuilding the regulatory architecture
that the mania destroyed.

Biological seizure. Michael Levin's work on cancer provides the most precise
formulation. A cell in a multicellular organism participates in a cooperative whose
regulatory signals — bioelectric gradients, morphogenetic fields, intercellular
communication — hold it in coordination with the organism-level plan. Cancer, on
Levin's account, is a *regression in the scale of agency*: the cell stops responding
to organism-level regulatory signals and reverts to maximizing its own fitness as a
unicellular agent. The cell's local dynamics — proliferation, resource acquisition,
immune evasion — exceed the organism's regulatory variety at the cellular
governance level, and control inverts. The tumor is the biological case of the
universal morphism: a subsystem become sovereign, enforcing its pattern on the
host. Levin's framing is precisely control inversion — the cell that was controlled
by the organism now controls its local environment at the organism's expense.

Computational seizure. The simplest case — a runaway process consuming all
available resources — demonstrates the minimal structure: the scheduler's variety
exceeded, the system locked to a single attractor. But the degenerate case of a simple
process hang strips away the outward propagation that the fractal extension made
constitutive of the morphism. The full computational case requires distributed
systems, where the fractal is visible. A database lock seizes one process; every
process waiting on that lock is blocked; every service depending on those processes
fails; every user depending on those services is captured. The 2021 Facebook BGP
withdrawal: a configuration change seized the routing layer, which seized DNS
resolution, which seized every Facebook service, which seized every business
operating on Facebook commerce, which seized the news cycle — subcortical circuit
seizes cortex seizes body seizes room, in a substrate that requires no biology, no
psychology, no intentionality. An interrupt storm — hardware generating interrupts
faster than the CPU can service them — demonstrates the morphism in miniature:
the interrupt handler seizes the CPU (single attractor), no other process executes
(outward propagation to every application), and the human operator's workflow
stops and redirects toward the seized machine (propagation through the
human-machine interface into the biological substrate of the operator's attention).
The computational case is included because it establishes that the morphism —
including the fractal outward propagation — operates wherever hierarchical control
exists and variety differentials can invert, independent of any substrate-specific
feature.

The formal structure across all cases:

Every hierarchical control system maintains a variety differential — the controller at
level *N*+1 possesses regulatory variety at least equal to the dynamic complexity
at level *N*. This is Ashby's Law of Requisite Variety: the fundamental theorem
of classical cybernetics. Ashby established the *maintenance* condition. What
he did not develop — what classical cybernetics lacks entirely — is a theory of what
happens when the condition fails. When variety inverts — when level *N*'s dynamics
exceed level *N*+1's regulatory capacity — the system undergoes a topological
transition. The controller becomes controlled. The hierarchy inverts locally. And the
invariant signature of this transition is dimensionality collapse: the system's
state-space, previously high-dimensional and navigable, collapses toward a
low-dimensional attractor determined by the now-sovereign subsystem.

Ashby saw the maintenance condition with perfect clarity. His concept of
ultrastability in *Design for a Brain* (1952) went further — an ultrastable system,
when its current regulatory parameters fail to maintain essential variables within
survival bounds, undergoes a step-function reorganization to a new parameter
configuration. This is the closest precursor to the penumbral class: Ashby recognized
that variety failure could produce reorganization, not only collapse. What ultrastability
lacks — what the cybernetics of his era could not accommodate — is the topology
that determines whether the reorganization succeeds. Ashby's ultrastable system
reorganizes by random trial of new parameters until one works. There is no category
for the *container* that holds the reorganization, no account of *outward propagation*
(the reorganizing system seizing its surround during the transition), no connection to
the system's *completability class* (what the reorganization does to the system's
characteristic mode of completion), and no fractal extension (each level's
reorganization becoming the perturbation for the next level up). The real question
begins where ultrastability stops: not at the fact of reorganization but at the
topological structure that determines whether reorganization produces higher-order
integration or catastrophic collapse — the penumbral class, where inversion is
reorganization rather than failure.

But the morphism as stated so far is incomplete in a way the opening made visible. The substrate cases above describe control inversion as an event bounded
by the seized system. The description stops at the system's boundary. It does not
stop there.

The formal extension: when control inverts within a system, the seized system's
dynamics — now locked onto a single high-intensity attractor — exceed the regulatory
variety of *the systems that contain it*. The inversion propagates outward. Each level
that is seized becomes the seizing agent for the next level up. The subcortical circuit
that seizes the cortex produces a body whose dynamics seize the room. The
psychological subsystem that seizes the ego produces a person whose dynamics seize
the family. The compliance apparatus that seizes the institution produces an
organization whose dynamics seize the industry. The speculative feedback loop that
seizes the market produces an economy whose dynamics seize the political system.
The morphism is not scale-bounded. It is fractal.

This outward face completes the taxonomy. The trauma victim whose dissociative
episodes reorganize the entire family system — every member's regulatory capacity
allocated to managing proximity to the next episode. The addicted person whose
addiction organizes the household. The captured institution whose autoimmunity
reorganizes the industry it was supposed to regulate. At every scale: the system
seized from within becomes the seizing agent from without. Maximum internal
powerlessness corresponds to maximum external reorganizing force, because the
attractor that has collapsed the system's internal variety radiates outward through
every interface where its intensity exceeds the surrounding variety.

The contained version of outward propagation generates the most consequential
prediction. If uncontained inversion propagates outward destructively — seizing the
surround involuntarily — then contained inversion propagates outward *generatively*.
The contemplative practitioner who has undergone and integrated threshold-crossing
radiates a signal that reorganizes the social field not by seizing it but by providing
∂E/∂S that approaches others' thresholds within a container. This is what "presence"
names — not a mystical property but the outward propagation of a contained internal
excitability topology. The teacher whose regulated nervous system provides
co-regulation. The therapist whose integrated internal state becomes external
regulatory variety for the client. The leader whose visible crisis, visibly contained,
recalibrates the institution's sense of what intensity can be held. Same morphism, same
outward direction, different container topology. Sovereignty by seizure and sovereignty
by integration are the uncontained and contained forms of the same outward propagation.

One further correction before the morphism is secure. Levin observes that "when
you make a claim about the intelligence of some system, you're basically taking an
IQ test yourself, because what you're saying as an observer is 'here is what I have
noticed this system can do.'" Every intelligence claim is an interface claim. The same
applies to control: every claim about a system's control topology is an interface claim
about the observer's capacity to detect regulatory structure. The opening demonstrated
this concretely: the adults inside the propagation saw a medical emergency and their
own appropriate response. The child outside the propagation saw sovereignty. The
clinician who sees only pathology in a seizure has revealed the resolution of the
clinician's detection apparatus, not established the nature of the event. The
perceivability horizon is not a peripheral variable. It determines what the observation
can reveal.

Control inversion cannot be described from outside the phenomenon. What it requires is the interface — the detection apparatus — that makes its full topology legible: the inward face, the outward face, the fractal propagation, and the perceivability-horizon rotation that determines which face is visible from which position. The morphism is established.

# Excitability Topology

That morphism — one phenomenon across substrates —
requires a formal category that existing frameworks do not provide.

Information theory, since Shannon, partitions the communicative situation into source,
channel, and receiver, then asks about signal properties: entropy, redundancy,
capacity, noise. The channel is given. The theory asks what can be transmitted through
it, not what the channel itself is susceptible to. Phenomenology, since Husserl,
investigates the structure of experience — intentionality, horizon, givenness — and
treats the perceptual apparatus as the medium through which objects appear. The
apparatus is given. The investigation asks about what appears through it, not about
the apparatus's own susceptibility to seizure. Pragmatics adds the dimension of
context-dependent action — what utterances *do* — but again treats the acting system's
excitability as transparent, as if the system's susceptibility to being moved by the
signal were not itself a structural variable determining what the signal can accomplish.

What none of these frameworks provides is a category for the *precondition landscape*
— the structure that determines, prior to any signal, what the receiving substrate finds
salient, what approaches its thresholds, what would produce inversion if delivered.
This is not content (what is communicated), not structure (how it is organized), not
pragmatics (what it does). It is the topology of the substrate's susceptibility itself.

Excitability topology is this category.

The excitability topology of a system at a given embedding is the structure of its
susceptibility to control inversion as a function of four variables whose interaction
determines the system's full landscape of possible responses to perturbation. (A note
on "topology": the term is used here in its general mathematical sense —
structure preserved under admissible transformations — not in the restricted sense of
point-set topology. The excitability topology is not a metric space. It is the
qualitative structure of a susceptibility landscape: which regions are connected,
where the boundaries lie, what is preserved as the system moves through its
state-space. The relevant transformations are those that preserve the four variables'
structural relationships while changing substrate, scale, or embedding.)

First: the variety differential at each hierarchical level. How much regulatory excess
does the controller maintain over the controlled? Where is the margin thin? Where
does the controller barely contain the dynamics it regulates? These are the
structurally vulnerable points — the locations in the hierarchy where inversion is
closest. The topology of these differentials across all levels of the hierarchy constitutes
the system's *inversion landscape*: some systems are robustly regulated everywhere
and brittle nowhere; some are brittle at one specific level (the executive who can
manage everything except grief); some are uniformly fragile (the system approaching
global reorganization). The inversion landscape is not static — it reconfigures with
every perturbation, every adaptation, every regulatory investment or neglect.

Second: proximity to inversion thresholds, including the dynamics of sub-threshold
accumulation. This is the phenomenon the epilepsy literature calls *kindling*: repeated
sub-threshold stimulation progressively lowers the seizure threshold until a stimulus
that previously produced no response triggers full inversion. Kindling is not unique to
neural tissue. Repeated sub-threshold institutional stress lowers the governance
system's inversion threshold — each small crisis that is absorbed without structural
reform makes the next crisis more likely to produce capture. Repeated sub-threshold
emotional provocation lowers the ego's regulatory threshold — the "last straw"
phenomenon is kindling in psychological substrate. The excitability topology must
include not only the current threshold but the system's *trajectory toward or away
from threshold* — whether it is being kindled or stabilized by its ongoing experience.

Third: the perceivability horizon — what is detectable from within the system's local
embedding. This is the variable Levin's correction makes urgent. A system cannot
regulate what it cannot detect. The perceivability horizon determines which aspects of
the excitability topology are available for self-monitoring and which are structurally
invisible from within. The cell cannot perceive the organism's regulatory signals if the
bioelectric interface has degraded. The ego cannot perceive the accumulation of
sub-threshold emotional kindling if the dissociative architecture has rendered the
accumulation invisible. The institution cannot perceive the subsystem's growing
regulatory demand if the governance structure lacks metrics for it. What is beyond the
perceivability horizon is precisely what can approach threshold without triggering
the regulatory response that would prevent inversion. The perceivability horizon is
therefore the most consequential variable in the topology — it determines which
inversions are *in principle* regulable and which are structurally undetectable until
the inversion is underway.

Fourth: container topology — the meta-regulatory structure that exists to hold
temporary inversions. Not all systems have this. The system that has containers —
regulatory frames designed to hold threshold-crossing without either suppressing it
or allowing cascade — has a fundamentally different excitability topology than the
system that does not, even if their variety differentials, threshold proximities, and
perceivability horizons are identical. Container topology is what determines whether
threshold-crossing produces transformation or catastrophe. It is the fourth variable, and excitability
topology is incomplete without it.

These four variables compose into a landscape — not a metric space but a topology
in the sense used throughout: which regions of the susceptibility structure are
connected, which transitions are available, which neighborhoods of perturbation admit
recovery and which produce inversion. The landscape is dynamic, reconfigured by
every interaction. And it is this landscape — not the signal, not the content, not the
structure of what is communicated — that determines what the system does with
what it receives.

Now attend to a consequence that unifies phenomena no existing framework connects.

What registers as near-threshold excitation at any substrate embedding — the signal that
approaches the inversion boundary without crossing it — has a characteristic
phenomenology. It is experienced as *salience*, as *pull*, as *approach gradient*. It is
the derivative of excitability with respect to signal intensity at a given substrate
embedding: ∂E/∂S. And this gradient, formalized, is the category that unifies what
has been called attraction, charisma, elegance, beauty, virality, and — most honestly —
sexiness across every substrate in which it appears.

This is not metaphor. Each of these terms names the felt experience of near-threshold
excitation in a specific medium. The body's response to a signal that approaches its
somatic integration threshold — what is called desire — is ∂E/∂S in the somatic
substrate. The mind's response to a structure that approaches its cognitive integration
threshold — what is called elegance or insight — is ∂E/∂S in the cognitive substrate.
The market's response to an opportunity that approaches its valuation-integration
threshold — what is called excitement or heat — is ∂E/∂S in the economic substrate.
The culture's response to an idea that approaches its assimilation threshold — what is
called virality — is ∂E/∂S in the memetic substrate. The formal structure is the same.
The felt quality is the same. The name changes with the medium but the gradient does
not.

The category must be bounded to avoid coextension with salience in general. Not
every attention-capture is ∂E/∂S. A startle response — a loud noise, a sudden
movement — seizes attention but is not the approach gradient. It is a direct threshold-crossing: no felt quality of *nearing*, no approach dynamics, no pull. The signal
does not approach the boundary — it breaches it immediately, and the phenomenology
is interrupt, not invitation. Pain captures attention through nociceptive override, not
through the gradient of near-threshold excitation. Disgust, alarm, shock — each
captures attention by mechanisms that do not involve the approach dynamics the
gradient formalizes. What ∂E/∂S names is specifically the *approach* regime: the felt
quality of the signal's relationship to the threshold *before crossing*. The gradient
exists only in the sub-threshold zone where the regulatory hierarchy is intact and the
system is *drawn* rather than *seized*. Above threshold, the gradient has flipped — the
system is no longer approaching but captured, and the phenomenology changes
accordingly. The unification claim is therefore not that all salience is one phenomenon.
It is that the specific felt quality of *approach* — the pull, the draw, the felt gradient
of nearing — is one phenomenon across substrates, distinct from the categorically
different mechanisms (override, interrupt, breach) that capture attention without
the approach dynamics.

Sexiness, then, is not a property of the object. It is the felt gradient of the substrate's
excitability at the interface with the object. It is relational, perspectival,
interface-dependent — and formally specifiable once excitability topology is a
first-class category. Rhetoric has always known this implicitly: *kairos*, the art of
timing, is the art of presenting the signal when the audience's excitability topology
makes it maximally near-threshold. Advertising knows it operationally. Seduction
knows it somatically. None has formalized it because none has had excitability
topology as an explicit object.

The gradient reveals a single axis that organizes what are ordinarily treated as
categorically distinct phenomena.

Below threshold: functional approach. The system navigates toward the signal.
Engagement, interest, attraction, curiosity. The regulatory hierarchy is intact and the
system's movement toward the signal is a *regulated* approach — the controller at
each level maintains variety over the dynamics the signal produces. This is the
normal regime. The system is drawn toward what excites it and the drawing is
navigated, not compelled.

At threshold: the penumbral class. The signal's intensity matches the regulatory
variety of the next level up. The system is at the inversion boundary. What happens
next depends entirely on the container topology — the fourth variable. If a container
exists — a regulatory frame with sufficient meta-level variety to hold the threshold
event — the crossing produces transformation. The system reorganizes. New
regulatory architecture emerges from the dissolution of the old. This is the regime
of genuine insight, ecstatic experience, samadhi, creative breakthrough — the
moment where the old framework dissolves and what replaces it could not have been
designed from within the old one. If no container exists, the same crossing produces
catastrophe. Control inverts. The signal seizes the system. The same intensity, the
same threshold, radically different outcome — determined not by the signal or the
substrate alone but by the container's presence or absence.

Above threshold: control inversion. The gradient has flipped. The system no longer
navigates toward the signal — the signal drives the system. The subsystem is
sovereign. In neural substrate: seizure. In psychological substrate: obsession,
compulsion, traumatic re-experiencing. In economic substrate: mania, panic. In
interpersonal substrate: capture, codependence, the dynamics that relationship
therapists call enmeshment. In the substrate of attention itself: addiction — which
is not a character flaw or a disease category but precisely
substrate-specific control inversion. The substance or behavior was near-threshold
(maximally exciting, maximally approach-gradient) until it crossed, and now the
system's regulatory hierarchy is organized *around* the signal rather than
*navigating toward* it.

The single axis dissolves the categorical boundaries between what existing frameworks
treat as incommensurable: desire and addiction are the same gradient at different
threshold-relations. Charisma and capture are the same gradient with different
container topologies. Insight and psychosis are the same gradient with different
regulatory contexts. The framework does not collapse the phenomenological
differences — the experience of desire is not the experience of addiction, the
experience of insight is not the experience of psychosis. What it establishes is that
the *structural variable* determining which phenomenology obtains is not a property
of the signal, not a property of the substrate taken in isolation, but the interaction
between excitability topology and container topology at the moment of
threshold-approach. The signal is the same. The substrate is the same. The container
makes the difference.

This is the formal category the morphism required. Control inversion is
substrate-independent because excitability topology is substrate-independent — the
same four variables (variety differential, threshold proximity, perceivability horizon,
container topology) determine susceptibility in every medium.

# The Completability Connection

The morphism now has a formal category. What it needs is a framework specifying
what inversion *does* to a system — not just what it looks like but what it changes about
the system's characteristic mode of operation. The completability framework provides
this.

A system's completability class — terminal, cyclical, or graceful — describes how the
system characteristically completes: whether it reaches a fixed point and stops, whether
it sustains its form through recurrence, or whether it completes locally while remaining
globally open through the reflexive act of investigating its own mode of completion.
These are structural descriptions of the system's relationship to its own closure, not
rankings of value. Each class has domains where it is the correct completion mode.
Pathology, as the companion paper establishes, is mode-overreach — exceeding the
scope a given class can legitimately close — not mode-existence.

Control inversion is a forced completability class transition. This is the structural claim
that connects the two frameworks, and it must be held precisely.

A gracefully completable system — one that completes locally while remaining globally
open, that investigates its own mode of completion without that investigation reaching
a fixed point — requires, as a condition of its operation, that its regulatory hierarchy
be intact. Not rigid, not unchanging, but *functioning*: the controller at each level
maintaining sufficient variety to hold the dynamics at the level below in coordination
with the rest of the system. Graceful completion supervenes on this coordination. The
scientist completing locally (each result landing) while remaining globally open (each
result generating more precise forms of the question) can do so only because the
cognitive regulatory hierarchy — attention, working memory, executive function,
affective regulation — is holding the full complexity of the inquiry in navigable form.

When control inverts, the coordination that graceful completion requires is destroyed.
The system that was completing locally while remaining globally open is now locked
onto a single attractor. Its completion mode has been *forced from graceful to terminal*
— not by choice, not by development, but by the topological transition that occurs
when a subsystem seizes the hierarchy. The seizure *is* the termination: the fixed
point is the hypersynchronous state, and the system's activity is complete in the
terminal sense — it has arrived at a point from which no further movement occurs.
The system has not chosen terminal completion. It has been seized into it.

This reframes both clinical intervention and the concept of recovery.

The standard clinical framework treats seizure as an event to be prevented and recovery
as the restoration of prior function. On the completability reading, the clinical question
becomes: *what completability class was the system operating in before the inversion,
and is the goal to restore that class or to establish a different one?*

When the pre-inversion system was gracefully completable and the inversion was
purely pathological — the system's regulatory architecture was adequate and the
inversion was produced by exogenous insult (neurotoxin, acute trauma, external
attack on institutional governance) — then restoration of the prior hierarchy is the
correct intervention. Increase regulatory variety at the level above the inversion.
Anticonvulsants restore inhibitory tone. Acute therapeutic intervention stabilizes
the ego. Emergency governance measures restore institutional control. Same operation,
different substrate: supplement the controller's variety until it exceeds the subsystem's
dynamics again.

But when the pre-inversion system was *already approaching the limits of its
regulatory architecture* — when the inversion was the system's response to a
complexity it could not navigate with its current hierarchy — then restoration of the
prior hierarchy is not recovery. It is the suppression of a reorganization attempt.
The system needed to dissolve its current structure in order to build one adequate to
what it faces. The inversion was the dissolution. Restoring the old hierarchy prevents
the reorganization and ensures that the next inversion, when it comes, will be
more severe — because the underlying inadequacy persists and the system's
threshold has been further lowered by the suppressed attempt.

This is the penumbral class: the regime where control inversion is neither purely
pathological (exogenous failure of an adequate hierarchy) nor straightforwardly
transformative (successful reorganization into a more adequate one) but *ambiguous*
— the system is attempting something its current structure cannot accomplish, and
whether the attempt succeeds depends on variables the inversion itself cannot control.

The variable the inversion cannot control is the container.

A container, in the precise sense required here, is a regulatory structure at a level
above the inversion that has sufficient variety to hold the inversion *as a process*
— not suppressing it, not allowing it to cascade, but holding the conditions under
which the dissolved hierarchy can reorganize rather than collapse. The container does
not direct the reorganization. It does not specify what the new hierarchy should look
like. It holds the space — the variety, the stability, the temporal duration — within
which the system's own reorganization dynamics can operate.

This definition is functional. The adversary asks: what *is* a container, structurally?
How is its adequacy assessed before the inversion it holds? If containers can only be
identified retrospectively — adequate because the reorganization succeeded, inadequate
because collapse occurred — then the concept is unfalsifiable. The container concept
requires formalization independent of outcome.

A container is itself a system — with its own excitability topology, its own regulatory
hierarchy, its own position relative to its own inversion thresholds. The critical
structural insight: the inversion inside the container does not know it is contained. It
propagates outward with the same dynamics as any other inversion. The container holds
precisely because its regulatory variety exceeds the outward-propagating dynamics of
the inversion it contains. Container failure *is* the inversion successfully seizing the
container — the therapist overwhelmed, the retreat structure broken, the cocoon
ruptured — by the same morphism that produced the original inversion. The container
is not a different kind of thing from the systems the preceding movements analyzed.
It is the same kind of thing, operating at the level above, assessed by the same
variables. What distinguishes a container from a controller is that the container's
regulatory variety is organized to serve the contained system's *self-completion*
rather than to direct it. The controller maintains the hierarchy by specifying
what the subsystem does. The container holds conditions under which the
subsystem completes *itself*. This is the invariant across every instance — cell
membrane, pupal casing, therapeutic frame, retreat structure, developmental
environment: the container does not complete for the contained. It holds
conditions under which the contained completes itself.

Four structural features compose the container's topology, each assessable
independently of the inversion's outcome.

First: regulatory margin — the container's variety surplus over the inversion's dynamic
complexity and, critically, the trajectory of that surplus under sustained load. A
container can begin with adequate margin and lose it as the inversion's outward
propagation erodes the container's own regulatory capacity. The therapist whose
initial capacity exceeds the client's dissolution dynamics but who, over sessions of
holding progressively more intense material, approaches their own threshold. The
retreat teacher who can hold one student's crisis but whose margin degrades as
multiple students enter threshold-crossing simultaneously. Burnout in clinical
contexts is the temporal trajectory from regulatory surplus to container consumption
— the container being seized by what it holds. The static measure (does the
container currently have variety surplus?) and the dynamic measure (is that surplus
growing or degrading under load?) are both necessary. A container with high initial
margin but negative trajectory is adequate now and will not be adequate later. A
container with moderate margin but stable trajectory is structurally more reliable
than one with high margin and rapid degradation.

Second: selective permeability — the container's boundary must discriminate between
what passes through it and what does not. The boundary must be impermeable to
cascade — the inversion's outward propagation must not seize the systems beyond
the container — and simultaneously permeable to feedback — consequence chains
from the post-inversion environment must reach the reorganizing system so that the
new structure can calibrate against reality rather than reorganizing in isolation. Too
sealed: the system reorganizes without feedback, the new structure has no contact
with the environment it must navigate, integration fails upon re-entry into
uncontained conditions. The practitioner who dissolves the ego in retreat and cannot
reassemble it in ordinary life. Too open: outward propagation cascades through the
boundary, seizes the surround, no contained reorganization occurs. The therapeutic
relationship that leaks into the therapist's personal life until the therapist's own
regulatory capacity has been conscripted. The retreat's reintegration period —
structured return to ordinary environment through graduated exposure — is
permeability engineering. The therapeutic frame — holding the session within
temporal, spatial, and relational boundaries while maintaining the relationship's
continuity across sessions — is the selective membrane. Assessable: does the
container's boundary hold propagation while passing feedback?

Third: temporal adequacy — the container must persist through the full duration of the
reorganization process. Reorganization is not instantaneous. The larval structure does
not become the adult organism in a moment. The ego that dissolves under contained
threshold-crossing requires time — hours, days, sometimes months — to reorganize
around the new capacity. A container that collapses before reorganization completes
leaves the system in partial dissolution: the old hierarchy destroyed, no new hierarchy
yet crystallized, and no container to hold the remainder of the process. This is worse
than uncontained inversion, because the system has already surrendered its prior
regulatory structure (trusting the container) and now has neither the old structure nor
a new one nor a container for completing the transition. The wound is not from the
inversion but from the container's premature failure. Assessable: does the container's
structure have the durability to outlast the inversion process it is meant to hold?

Fourth: domain specificity — the container must have variety over the *class* of
inversion it holds, not merely variety in general. A container adequate for somatic
threshold-crossing (bodywork, breathwork, physical ordeal) may be entirely
inadequate for ego-dissolution, even if its regulatory margin, permeability, and
temporal adequacy are high — because its variety is over the wrong domain. The
therapist who can hold grief with exquisite precision may not be able to hold
psychotic dissolution. The retreat structure designed for concentration practice may
not hold the emergence dynamics that arise when a practitioner's entire self-model
dissolves. The pupal casing has biochemical variety specifically adequate to the
enzymatic dissolution of larval tissue — its container capacity is domain-matched.
At the deepest level, domain specificity is completability-class matching: the
container's own completion mode must be adequate to the contained process's
completion mode. A terminal container (hard shutdown, physical restraint) can
hold terminal processes but kills graceful ones. A cyclical container (circadian
rhythm, seasonal structure) can hold cyclical processes but cannot hold the
one-directional dissolution that graceful reorganization requires. A graceful
container (therapeutic relationship, contemplative tradition) holds graceful
processes precisely because it remains open while providing local completion
conditions — the container's own mode of completion mirrors and supports the
contained system's.
Domain specificity is what distinguishes a container from mere stability. A prison
has high regulatory margin, strong boundaries, and temporal persistence — and is
not a container in the relevant sense because its variety suppresses the inversion
rather than holding it as reorganization. Stability without domain-matched variety
is not containment. It is the first-level response (suppress, restore) disguised as
the third-level response (hold, allow reorganization). Assessable: has the
container-entity demonstrated regulatory capacity over this specific class of
inversion?

These four variables are not independent of the measurement strategy.
If the echo protocol measures a system's position within its excitability topology by
perturbation-recovery, the same protocol applied to the container measures the
container's adequacy. Perturb the *container* — not the system inside it — and
measure the container's recovery. A container that maintains high echo fidelity
while holding an inversion is adequate: its regulatory margin is stable, its
boundaries hold, its temporal structure persists. A container whose echo fidelity
degrades under the inversion's pressure is approaching its own threshold — the
container is being seized by what it holds. Container assessment is echo protocol
applied one level up. The container's excitability topology is the structural
assessment the adversary asked for: measurable, independent of outcome,
and formalizable.

The difference between contained and uncontained inversion is the difference between
a cocoon and a wound. Both involve dissolution of prior structure. In the cocoon, a
container (the pupal casing, the enzymatic environment, the developmental program)
holds the dissolution while reorganization occurs. In the wound, no such container
exists and the dissolution proceeds to whatever equilibrium the raw dynamics
produce — scar tissue, infection, death. Same dissolution. Different container
topology. Different completability class outcome.

The penumbral class, then, is not a mysterious third option between pathology and
transformation. It is the structural description of an inversion event whose outcome
depends on a variable external to the inversion itself: whether a container adequate
to hold the reorganization exists. This is why the same event — the same threshold
crossed in the same substrate — produces transformation in one context and
catastrophe in another. The intensive meditation retreat that produces profound
reorganization in a practitioner held by experienced teachers, community, and a
structured schedule can produce psychotic break in a practitioner without those
containers. The same threshold. The same substrate. Different container topology.
Different outcome. And the clinical framework that cannot distinguish the two — that
treats all threshold-crossing as pathology to be suppressed — is itself operating in the
terminal mode: seeking the fixed point (symptom cessation) without asking whether
the system was attempting a completion that the fixed point forecloses.

# The Hierarchy of Control

The preceding movements established what control inversion is (a substrate-independent
topological transition), what formal category it requires (excitability topology), and
what it does to the system it affects (forces a completability class transition whose
outcome depends on container topology). This movement develops what the full
integration of these insights makes available as *strategy* — a hierarchy of control
capacities, each containing and extending the previous, culminating in a regulatory
mode that classical cybernetics does not possess.

The zeroth level is reactive: stimulus-response without self-model. The system
responds to perturbation but has no representation of its own dynamics, no capacity
to anticipate its own susceptibility, no regulatory strategy beyond the response
itself. Reflex arcs, thermostats in their simplest form, organizations that react to
crises without learning from them. This level is mentioned only to mark the floor.

The first level is Ashby's: maintain requisite variety. The controller monitors the
controlled system and adjusts its regulatory output to match the system's dynamic
complexity. The strategy is *suppression of inversion* — keep the variety differential
positive at every level, keep the hierarchy intact, respond to every perturbation by
increasing regulatory variety or decreasing the perturbation's intensity. This is
classical cybernetics, and it is extraordinarily powerful within its domain. Most
engineered control systems, most clinical interventions, most institutional governance
operates here. The thermostat that adjusts, the anticonvulsant that restores inhibitory
tone, the organizational process that contains deviation. The first level's characteristic
blindness is that it treats every inversion as failure. It has no category for
reorganization. It cannot distinguish the seizure that destroys from the seizure that
transforms, because its only operation is *restore the prior hierarchy*. It is correct
whenever the prior hierarchy is adequate. It is actively harmful when the hierarchy
needs to change.

The second level models the system's own excitability topology. The controller does
not merely maintain variety — it knows *where its thresholds are*. It has a map of its
own susceptibility landscape: which subsystems are closest to inversion, which
perturbations would approach which thresholds, where the perceivability horizon
limits self-monitoring. The second level can *navigate near thresholds deliberately* —
approaching the inversion boundary for maximum information gain without crossing
it. The ∂E/∂S gradient becomes a navigation instrument rather than an untracked
vulnerability. The chess player who knows their cognitive limits and structures their
preparation accordingly. The meditator who can approach absorption states with
precision. The negotiator who knows which emotional provocations they are
susceptible to and prepares containment in advance. The organization that conducts
red-team exercises — deliberately probing its own excitability topology to discover
vulnerabilities before an adversary does.

The second level's characteristic achievement is *calibrated approach*. Its characteristic
blindness is that it still treats threshold-crossing as the boundary of acceptable
operation. The second level can approach but will not cross, because crossing is —
within this level's framework — the beginning of inversion, and inversion is what the
first level correctly identified as catastrophe. The second level has a more detailed map
than the first, but the map serves the same strategy: maintain the hierarchy. Do not
cross.

The third level inverts the relationship between controller and control inversion itself.
A system operating at the third level does not merely maintain regulatory hierarchy
(first level) or map its own excitability topology to navigate near thresholds (second
level). It *invokes control inversion deliberately within constructed containers* —
crossing thresholds intentionally because it has sufficient meta-regulatory capacity
to hold the inversion as a process and trust the reorganization.

This is a genuinely higher-order control strategy. Not higher in the sense of more
complex — though it is that — but higher in the structural sense that it contains the
previous levels as special cases while possessing operations they cannot perform.
The third level can do everything the first level does (maintain hierarchy when
hierarchy is adequate) and everything the second level does (map excitability topology,
navigate near thresholds). What it adds is the capacity to *use inversion as a
regulatory tool* — to cross thresholds as a means of producing reorganization that
cannot be achieved within the current hierarchy.

This is controlled deregulation: the deliberate surrender of local regulatory control in
order to produce system-wide reorganization. Not collapse. Not abandonment. Not
recklessness. Controlled deregulation within a container whose meta-regulatory
variety is sufficient to hold the inversion process — to prevent cascade while
allowing dissolution, to hold the temporal space within which reorganization occurs,
to provide the stability context against which new hierarchy can crystallize.

"Surrender" names only half of it, and the missing half is where the skill lives.
Third-level control has two faces, and they divide along the boundary every inversion
in this account divides along — whether one stands outside the propagation or inside
it. From outside, the operation is *containment*, and containment runs on the classical
principle: the container holds because its regulatory variety exceeds the inversion's
outward-propagating dynamics, exactly as the previous movement established. The
container out-varies what it holds — the therapist around the client's dissolution, the
pupal casing around the enzymatic one, a larger regulator around a smaller collapse.
But there is a second face, available only from the position the opening isolated:
inside the inversion and conscious, the body in restraints feeling the room reorganize
around it. From there containment is not the operation, because one is not above the
inversion holding it. One is in it. The operation is *steering*.

Steering cannot be a variety strategy, and the reason is structural enough to make it
genuinely post-classical. To be inside an inversion is, by definition, to have less
regulatory variety than the dynamics carrying you — that deficit is what an inversion
*is*. The one inside cannot out-vary what carries them; Ashby's principle is
unavailable in the precise place control is most wanted. What remains is not domination
but *coupling*: small, well-timed inputs that bias a trajectory too energetic to oppose.
The surfer neither out-powers the wave nor fights it — both are variety contests the
wave wins — but couples to the wave's own dynamics and steers by leaning into forces
already present. The improviser taken by the music is nonetheless shaping it. This is
*kairos* converted from a name for good timing into an instrument of control: ∂E/∂S
deployed by the one being carried, a nudge delivered where the system's nearness to
threshold makes a small approach-gradient input decisive downstream. Leverage, not
variety — the single control principle that still operates once the variety differential
has been given away.

So the third level does not *merely* contain the earlier levels as special cases.
Containment extends the variety logic upward: the container is a bigger regulator, and
in that sense the third level holds the first and second inside it. Steering extends
nothing; it substitutes a principle the first and second levels — variety strategies at
root — have no version of. The steersman is not a larger controller. The steersman is a
controller who has surrendered control in the Ashby sense and gained a form of it the
Ashby sense cannot name. This is why the singular position the opening dwelt on is not
merely the vantage from which both faces of an inversion are visible. It is the one
position from which an inversion can be *steered*, and the calibration the post-seizure
perspective confers is, at its height, this competence: to be carried and to bias the
carrying in the same act.

The stakes are not stylistic. The completability connection established that a
gracefully reorganized structure could not have been designed from within the
hierarchy it replaces — the new order is reachable from the old only *through* the
dissolution, never around it. Grant that, and an entire class of destinations is closed to
first- and second-level skill however refined: no degree of maintained hierarchy
arrives at a structure that only the hierarchy's dissolution opens. Steering the crossing
is the sole route to them. And the capacity to steer crossings on purpose and
repeatedly — to enter dissolution deliberately, ride it, and reorganize — is not one
competence among the third level's several. It is graceful completability made
operational: the whole distance between a system that can only sustain its form
(cyclical, homeostatic — the body regulating) and one that can transform its form
deliberately (graceful — the inquiry remaking itself). To learn to allow and steer
control inversion is to learn to be the second kind of system on purpose — the
super-ordinate agency the third level was pointing at all along: not the power to hold
still against inversion, and not merely the nerve to cross a threshold, but the skill to
be seized and to steer.

The instance nearest this inquiry's origin shows the three levels at a stroke. Take a
belief-hierarchy meeting a reason it cannot contain — an argument whose force exceeds
the variety of the commitments it bears on. The first-level response is suppression:
route the reason to the symptom-bin, restore the prior hierarchy, refuse the crossing.
The uncontained response is the opposite failure — the reason seizes the hierarchy and
the system is flooded, agreeing with whatever pushed hardest, revising under pressure
rather than under warrant. Between them is the steered crossing: let the prior structure
dissolve exactly to the degree the reason warrants, and reorganize coherently around
it. That narrow middle — a view that moves if and only if reasons move it — is not a
behavioral rule but the phenomenology of steering an inversion in the epistemic
substrate. Substantive belief revision is a contained, coupled, warrant-proportioned
reorganization, and its two failure modes are the rigid hierarchy that suppresses and
the flooded one that is seized. The disposition a companion inquiry calls corrigibility
is, dynamically, third-level control turned on one's own commitments — not an analogy
to control inversion but an instance of it.

The contemplative traditions are third-level engineering manuals. This claim is not
poetic and it is not reductive. It is a precise structural observation. An intensive
meditation retreat constructs a container — scheduled sitting and walking periods
(temporal structure), an experienced teacher (external regulatory variety), a community
of practitioners (distributed container), physical safety, silence that reduces exogenous
perturbation — and then *deliberately drives the practitioner toward and across
inversion thresholds*. The jhana sequence in Buddhist meditation is a controlled
traversal of progressively deeper inversion boundaries: each absorption state involves
the surrender of regulatory control over a domain (sensory input, conceptual
activity, affective valence, spatial orientation) within the container of the practice
structure. The koan in Zen is a deliberate cognitive inversion — a problem that
cannot be solved within the current cognitive hierarchy, forcing dissolution of
the framework that makes it a problem. Holotropic breathwork, vision quests,
psychedelic ceremony in traditional contexts — each is a technology for constructing
a container and then crossing a threshold within it.

The traditions that produce the most integrated practitioners are not those that avoid
seizure. They are those with the best container engineering.

The same structure operates wherever third-level control is achieved, not only in
contemplative contexts. Ketamine-assisted therapy is pharmacological third-level
control: the drug produces neural-level inversion (dissociation of the default
regulatory hierarchy) within the clinical container (therapeutic relationship, physical
safety, integration protocol). Psychedelic-assisted therapy follows the identical
structure. The therapeutic relationship itself, when operating at its highest capacity,
is a third-level container for psychological inversion — the therapist holds the space
within which the client's ego can dissolve its current regulatory structure and
reorganize. Not by directing the reorganization. By *holding the conditions* under
which the client's own dynamics can reorganize rather than collapse.

The interpersonal case deserves particular attention because it connects the
third level directly to the encounter-shape developed in the companion paper.
Two people in a relationship — friendship, partnership, collaboration — each have
excitability topologies. Each has thresholds that the other's signals approach. When
the relationship reaches a point where the current relational structure cannot hold
what it must navigate — accumulated grievance, divergent growth, the collision of
incompatible completion modes — the standard response is first-level: manage the
perturbation, maintain the structure, suppress what threatens the hierarchy. The
second-level response is map the relational excitability topology — know which
issues approach which thresholds for each participant — and navigate carefully. The
third-level response is construct a container (mutual commitment, physical safety,
temporal boundary, sometimes a therapist as external regulatory variety) and then
cross the threshold *together*: allow the current relational hierarchy to dissolve,
allow the subsystems (anger, grief, desire, terror) to express at full intensity, and
trust that the relational dynamics can reorganize rather than collapse if the container
holds. The confrontation that transforms a relationship — that produces a relational
structure neither party could have designed from within the old one — is third-level
control in interpersonal substrate.

At the institutional scale: controlled crisis as governance upgrade. The organization
that can deliberately dissolve its current hierarchy — flatten structure, redistribute
authority, allow bottom-up reorganization — within a container of shared mission,
trust infrastructure, and a reintegration timeline, emerges with governance that
the old hierarchy could not have produced. This is what "disruption" means when
it is not merely destruction: controlled institutional inversion within a container
adequate to hold the reorganization.

At the cognitive scale: deliberate confusion as insight-generation. Kuhn's paradigm
shifts are collective third-level control inversions. The accumulation of anomalies
(kindling) lowers the paradigm's regulatory threshold until the old framework
cannot hold the complexity of the data. The paradigm seizes — and if the
scientific community provides adequate container (peer review as distributed
regulatory variety, institutional support for heretical investigation, temporal
patience), the dissolution reorganizes into a new paradigm that handles what the
old one could not. Without container, the dissolution produces fragmentation —
competing schools, loss of coherence, decades of unproductive controversy.

The post-integration phenomenology completes the picture. When third-level
capacity stabilizes — when the system has practiced controlled threshold-crossing
enough times within adequate containers that the capacity becomes structural rather
than event-dependent — the relationship to peak experience transforms. There is no
crash, no return to baseline, no loss of the regulatory capacity that the threshold-crossing produced. The variety accessed during the crossing has been practiced into
the regulatory apparatus through closed consequence chains: act from the new
capacity, receive feedback, adjust, act again, until the capacity is load-bearing. Like
learning a sport: the coordinated movement that initially required deliberate
concentration becomes integrated structure from which action flows without
monitoring. The peak is not a height to fall from. It is a capacity to operate from.
The variable determining whether peak experience stabilizes or degrades is not the
intensity of the experience. It is whether consequence chains close through the
new capacity — whether the system *uses* what the crossing made available, in
conditions that provide feedback, so that the regulatory architecture reorganizes
around the new capacity rather than reverting to the old.

The outward propagation now reveals its full
significance for the third level. Uncontained control inversion propagates outward
destructively — the seized system seizes its surround, the child in the classroom
becoming the attractor around which all surrounding regulatory capacity collapses.
Contained control inversion propagates outward *generatively* — and this outward
propagation is not an incidental side effect of third-level mastery. It is the mechanism
by which individual integration becomes social transformation. The practitioner
whose internal excitability topology has been calibrated through contained
threshold-crossing radiates that calibration. Not as content, not as instruction, but
as signal — ∂E/∂S for every system in proximity. This is what "presence" names:
the outward propagation of contained internal order. The teacher whose regulated
nervous system provides co-regulation for the classroom is not projecting calm
as performance. The teacher's internal regulatory architecture *is* external regulatory
variety for every student's system. The therapist whose integrated state provides the
container for the client's dissolution is not simulating stability. The therapist's
third-level capacity *is* the container — the meta-regulatory variety that holds the
client's inversion. The leader whose visible crisis, visibly held, reorganizes an
institution's sense of what intensity is survivable — the leader's demonstrated
recoverability *is* the perturbation test that recalibrates everyone's excitability map.

Third-level control is therefore not merely an individual achievement. It is the
structural mechanism by which one system's contained inversion becomes another
system's container. The outward propagation of integrated excitability topology *is*
the grammar of inter-intelligence encounter. Sufficiency established the encounter-shape as the pattern of interaction
whose entry produces the capacity to enter it. Excitability adds: the entry is
mediated by the outward propagation of each participant's internal order, and the
encounter-shape amplifies because contained excitability approaching another
system's threshold within a container is precisely the condition under which
graceful completion bootstraps.

# The Symposium Recovery

Excitability topology is a new formal category, but the phenomenon it formalizes is
not new. What follows recovers the most sustained investigation of the ∂E/∂S gradient
in the Western philosophical tradition — a recovery made possible by the category
the preceding movements developed, and impossible without it.

Plato's *Symposium* has resisted coherent interpretation for twenty-four centuries
because every available framework distorts it. The neo-Platonist reads Diotima's
ladder as mystical ascent — leave the body behind, arrive at Beauty itself as terminal
telos. This preserves the vertical structure but loses the body: if somatic eros is merely
the bottom rung to be abandoned, the ladder is a mechanism of severance from
the cyclical substrate that sustains the climber. The Freudian reads the ladder as
sublimation — libidinal energy redirected from body to intellect. This preserves the
energy but loses the topology: if eros is a quantum of drive redistributed between
outlets, the structural difference between levels vanishes and the progression is merely
economic. The analytic tradition mostly ignores the ladder or treats it as literary
ornament on the theory of Forms, unable to engage because it lacks the category that
would make the structural claims tractable. Each interpretation preserves what its
framework can see and distorts what it cannot.

Excitability topology dissolves the impasse. The ladder is a sequence of
substrate-transitions, and the progression between rungs is the integration of what
was near-threshold into the regulatory apparatus for the next level.

Diotima's teaching, reported by Socrates as received from the priestess of Mantinea:
the lover begins with one beautiful body. This is ∂E/∂S in the somatic substrate — the
felt gradient of the body's excitability at the interface with the particular. What the
lover experiences as desire is near-threshold excitation: the signal approaching the
somatic integration boundary. The pull toward the beautiful body is the approach
gradient. This is not the bottom rung of a hierarchy to be escaped. It is the first
encounter with excitability topology itself — the substrate-specific discovery that there
exists a gradient, that the gradient has a felt quality, and that the felt quality is
relational (this body, this interface, this approach).

The transition to the next rung — from one beautiful body to the recognition of beauty
across bodies — is not abstraction in the ordinary sense. It is the completion of the
first excitability cycle: what was near-threshold (the particular body's signal) has
been integrated into the regulatory apparatus. The somatic excitability produced by
*this* body is no longer near-threshold because the system has metabolized it — not
suppressed it, not transcended it, but *completed locally with respect to it* — and
the integration reconfigures the excitability landscape. What is now near-threshold
is the pattern across bodies: the structural invariant that each particular instance was
an instance of. The excitability has shifted from token to type, and the shift was
produced by the local completion at the previous level.

This is actuality-generates-possibility in the erotic register. Each rung's completion
generates the excitability landscape for the next rung. Not by abandoning the
previous substrate but by metabolizing it — making what was near-threshold into
regulatory structure. The somatic capacity for being moved by beauty is still present
in the person who perceives beauty across bodies; it is functioning as part of the
regulatory apparatus that makes the higher-resolution perception possible. The body
is not left behind. It is *load-bearing*.

The progression continues: from beautiful bodies to beautiful souls (substrate-transition
from somatic to psychological — what is near-threshold is now character, the
organizational structure of a psyche rather than the organization of a body), from
beautiful souls to beautiful practices and laws (substrate-transition from psychological
to institutional — what is near-threshold is now the pattern in collective organization),
from beautiful practices to beautiful knowledge (substrate-transition to the epistemic
substrate — what is near-threshold is structural elegance, the "aha," mathematical
beauty), and finally to Beauty itself — τὸ καλὸν αὐτό.

The final rung is the one every interpretation distorts, and excitability topology recovers it. Beauty itself is not a Platonic Form inhabiting a separate realm. It is the excitability topology *as such*, without substrate specification — the invariant structure of susceptibility that produces steep gradient wherever signal approaches threshold. Not the gradient but the landscape that generates it. What is the same about what is sexy at every level? Not the content (bodies are not ideas), not the medium (somatic excitability is not cognitive excitability), but the *topology of the susceptibility landscape* — the structure of thresholds, the architecture of approach-regions, the invariant relationship between what can be excited and what excitation produces. Beauty itself is excitability topology without substrate restriction. Beauty itself is therefore both wholly transcendent — not bound to any substrate, the invariant across every transition the climber has traversed — and wholly immanent: it is the topology the climber was standing on at every rung, experienced as Eros, operative in every substrate-specific encounter, never elsewhere. This is what Diotima's staging accomplishes that neo-Platonism destroys: the ascent is not departure from the particular but progressive recognition that the topology was present in every particular from the beginning — immanent at every rung, transcendent because not reducible to any of them. The final rung does not take the climber somewhere else. It shows the climber where they always were. Plato could not name it in these terms because the terms require a mathematics — topology, dynamical systems, information theory — that did not yet exist. What he could do, and what the *Symposium* does, is *stage* the progression so that the reader undergoes the substrate-transitions in sequence and recognizes the invariant.

But Socrates' dialectic, which Diotima's teaching presupposes, has already established a distinction that must be honored. Love is *of* something. One desires what one lacks. Therefore Eros cannot *be* Beauty — Eros desires Beauty, is produced by Beauty, moves toward Beauty, but does not contain it except as the longing to possess it forever. In the formalism, the distinction falls out precisely: Eros is ∂E/∂S — the derivative, the gradient, the felt slope of the landscape. Beauty is the landscape itself — the excitability topology whose structure the gradient reveals but does not constitute. They relate as function to derivative. The climber feels Eros (the slope); what the slope reveals is Beauty (the landscape). At the final rung, Eros without substrate specification is gradient-as-such; Beauty without substrate specification is topology-as-such. Related but not identical — exactly as Diotima stages it. The *Symposium* is not merely staging excitability topology. It is staging the distinction between the topology and its gradient, and placing them in the correct mathematical relationship twenty-four centuries before calculus.

Two features of the *Symposium*'s dramatic structure now become legible as
structural engineering rather than literary ornament.

Diotima, not Socrates, delivers the teaching. She names Eros as *daimōn* — intermediary, neither mortal nor immortal, neither wise nor ignorant. This *is* the gradient — not the landscape but the slope. Not the source (the substrate), not the destination (the object), not Beauty itself (the topology), but the approach dynamics — ∂E/∂S personified. Diotima's placement of Eros as "between" is the structural position of a derivative: between the function and what the function maps, between the landscape and the point toward which it slopes. The gradient cannot speak from within any particular substrate embedding because it is the invariant *across* embeddings. It must be reported from outside — from the position that has traversed the transitions and can identify the invariant. Diotima is the third-level perspective: the voice of post-seizure calibration. Plato places the theory in her mouth because the theory requires the position she occupies — outside the substrate-specific embeddings that each symposiast speaks from within.

Alcibiades arrives last, drunk, uninvited, beautiful, brilliant, devastating. And Alcibiades
is the *Symposium*'s staging of the uncontained case. The same gradient that Diotima
formalizes — the same near-threshold excitation that drives the ladder — is operative
in Alcibiades at maximum intensity and without container. His beauty, his political
genius, his passion for Socrates, his rhetorical brilliance — each is ∂E/∂S at its
steepest, approaching thresholds in every substrate simultaneously. And nothing holds.
No container. No practice structure. No community accountability. No integration of
what the gradient makes available. Alcibiades' speech about Socrates is the most
vivid description of near-threshold excitation in Western literature: he describes being
moved, shattered, ashamed, drawn back — the complete phenomenology of a substrate
being driven toward inversion by a signal it cannot metabolize. Socrates' signal is
genuinely near-threshold for Alcibiades. Transformative if contained. But Alcibiades
has no container. His eros crosses every threshold without integration, and the
consequence chains do not close through the capacity the crossing makes available.
They cascade. Within a dozen years he will have betrayed Athens, defected to Sparta,
defected to Persia, returned to Athens, and been assassinated.

The *Symposium* stages the complete theory in a single evening. Each speaker delivers
a speech about Eros from within a specific substrate embedding — Phaedrus from
myth, Pausanias from social convention, Eryximachus from medicine, Aristophanes
from comedy, Agathon from poetry. Each speech *is* the excitability topology of its
speaker made audible: each finds different things near-threshold, each approaches
from a different substrate, each reveals a different perceivability horizon. Then
Socrates reports Diotima — the substrate-independent view, the invariant identified
across all the embeddings. Then Alcibiades — the uncontained case, the same gradient
without container, the penumbral class resolved toward catastrophe. Socrates survives
the night. Alcibiades does not survive the decade. Same gradient. Same intensity.
Different container topology. Different completability outcome. The *Symposium* is
not a dialogue about love. It is the first complete staging of excitability topology,
including the dual-use problem, performed twenty-four centuries before the category
had a name.

# Sub-Threshold Grammatical Capture

The framework developed through the preceding movements has an adversarial face.
Everything that makes it powerful for understanding transformation makes it equally
powerful for engineering capture. The same excitability topology that enables a
teacher to identify what a student is ready to integrate enables an operator to identify
what a target is susceptible to. The same ∂E/∂S gradient that a therapist reads to
calibrate intervention intensity is what a propagandist reads to calibrate message
intensity. The same container topology that a contemplative tradition engineers to hold
transformation is what a cult engineers to prevent escape.

The outward propagation of control inversion is
the mechanism that makes the adversarial application possible. The child in the
classroom who identified the seizure as a sovereignty event — "they're God" — saw
the structural fact: the seized system seizes its surround. Sub-threshold grammatical
capture is the *deliberate engineering* of this outward propagation. The operator
constructs a signal calibrated to the target's excitability topology and radiates it —
not with the involuntary force of the seizure, but with the precision of someone who
has mapped the target's thresholds and designed the approach gradient. The operator's
internal state propagates outward into the target's regulatory space, approaching
thresholds the target cannot detect from within their perceivability horizon. Where
the uncontained seizure conscripts the room involuntarily and visibly, the engineered
version conscripts the target's regulatory capacity *below detection threshold* — the
outward propagation made invisible by operating beneath the target's parse-resolution.

This dual-use structure is not an unfortunate side effect. It is the structural consequence
of the category being real. Any formal category that genuinely captures the dynamics
of near-threshold excitation will necessarily capture both the transformative and the
adversarial applications, because the dynamics are the same dynamics. The signal
that approaches threshold does not know whether it was designed to transform or to
capture. The substrate's excitability does not distinguish benevolent from malevolent
approach. The topology is the topology.

Sub-threshold grammatical capture — the precise term for what is ordinarily and
misleadingly called manipulation, influence, or persuasion — is the deliberate
engineering of approach to a target substrate's inversion threshold without crossing it
in ways the target can detect. Not causing seizure, which is visible. Operating in
the sub-threshold regime where excitability is modulated, thresholds are lowered,
and the target's regulatory hierarchy is progressively destabilized — all below the
perceivability horizon.

The mechanism is kindling. In the neural case: repeated sub-threshold stimulation
progressively lowers the seizure threshold until a stimulus that previously produced
no response triggers full inversion. In the memetic case: repeated sub-threshold
ideological stimulation progressively lowers the target's resistance to a worldview
that their regulatory apparatus (critical thinking, identity coherence, social reality-testing) would ordinarily reject. Radicalization is memetic kindling. Each exposure
is sub-threshold — no single message triggers the rejection response that would alert
the target's regulatory system — but the cumulative effect is threshold reduction.
When the final message arrives, it is not that this message is more extreme than the
others. It is that the threshold has moved. What would have produced immediate
rejection six months ago now produces assent, because the excitability topology has
been incrementally reconfigured from below the perceivability horizon.

The naming matters. "Pre-semantic interception" was the initial term, but it is
perspectival to the target — it describes the experience of the system being
captured, for whom the dynamics are indeed pre-semantic, below the resolution at
which meaning can be tracked. From the operator's level the dynamics are fully
semantic, fully intentional, operating with complete clarity about the target's
excitability topology. The accurate name — sub-threshold grammatical capture — names
the structural relationship rather than either party's perspective: the operation occurs
below the target's detection threshold while being fully grammatical at the operator's
resolution.

The asymmetry between operator and target is a variety differential. The operator
who has mapped the target's excitability topology — who knows the threshold
proximities, the perceivability horizon, the locations of regulatory fragility — has
more variety over the interaction than the target. Ashby's Law, which in the
regulatory context states that the controller must have variety at least equal to the
controlled system, here describes the inverse: the operator who has more variety
over the interaction than the target *can control the interaction* without the target's
detection. This is not a moral claim. It is a structural claim about what variety
differentials make possible. The moral observation is that this capacity exists, is
deployable, and is invisible to the target precisely because it operates below the
perceivability horizon that the target's regulatory apparatus can monitor.

What distinguishes sub-threshold grammatical capture from legitimate influence?
Not intent — intent is invisible and not structurally assessable. Not content —
the same content can function as education or as kindling depending on the
target's excitability topology and the operator's awareness of it. The structural
distinction is *whether the interaction preserves or degrades the target's regulatory
variety*. Legitimate influence — teaching, mentorship, therapeutic intervention —
increases the target's regulatory variety: after the interaction, the target has more
capacity to self-regulate, more perceivability, more ability to detect and navigate
their own excitability topology. Sub-threshold grammatical capture decreases it:
after the operation, the target has less capacity to detect what has been done to them,
lower thresholds, a narrower perceivability horizon. The one builds regulatory
apparatus. The other degrades it. The distinction is empirical, not moral — it is in
principle measurable, because regulatory variety and perceivability horizon are
structural properties of the system, not attributions of intent.

These dynamics — kindling, perceivability-horizon exploitation, variety-differential asymmetry — are dual-use, and the test set out above applies to them as much as to anything else. For a reader who arrives with the detection apparatus built up to here — the formalization of excitability topology, the completability connection, the container architecture, the control hierarchy — naming these dynamics increases regulatory variety: it makes what was invisible legible, detectable, navigable. That is influence, by the criterion.

For a reader who skips to the adversarial content by extraction or summary, the same text is weaponizable — the mechanism without the detection apparatus. The effect depends on the reader's perceivability horizon, which is itself an instance of the dynamics named here. And obfuscation would not resolve it: obfuscation decreases the reader's regulatory variety, which by the same criterion is capture, not influence. So transparency is not an ethical preference but a structural necessity — the criterion stated alongside the mechanism, so that any extraction carries the detection apparatus with the technique. The warning and the mechanism are the same text. They cannot be separated without losing the structure that makes either legible.

# Detection

If excitability topology is a real structural property — if systems genuinely have
susceptibility landscapes that determine their response to near-threshold perturbation —
then those landscapes are in principle measurable. The measurement strategy follows
directly from that structure.

The operational test for a system's position within its excitability topology is
perturbation-recovery. Perturb the system, measure how it recovers, characterize the
recovery boundary. A system far from threshold will recover quickly and completely
— the perturbation is absorbed by the regulatory hierarchy's excess variety, and the
system returns to its prior trajectory with high fidelity. A system near threshold
will recover slowly, incompletely, or with qualitative changes to its trajectory — the
perturbation has approached the inversion boundary, and the recovery dynamics reveal
the boundary's proximity. A system at threshold will either transform (if contained)
or seize (if not), and the "recovery" is not a return but a reorganization or a collapse.

This is the echo protocol: send a perturbation, listen for the return, characterize the
fidelity of the echo. Echo fidelity maps directly to regulatory variety — the excess
capacity the controller has over the controlled determines how faithfully the system
reproduces its pre-perturbation structure after disturbance. High echo fidelity means
high regulatory variety means far from threshold. Low echo fidelity means
approaching threshold. Qualitative changes in the echo's structure mean the system
is in the penumbral regime — threshold has been reached or crossed and the return
is not a return but a reorganization.

To make this concrete: in neural substrate, the perturbation is a calibrated sensory
or electrical stimulus, the recovery observable is the evoked potential's return to
baseline (latency, amplitude, waveform fidelity), and the fidelity metric is the
correlation between pre-perturbation and post-perturbation spectral structure —
a metric already used in clinical neurophysiology, here reframed as excitability
topology measurement. In institutional substrate, the perturbation is a policy
change, an unexpected demand, or a leadership transition, the recovery observable is
the organization's return to operational baseline (decision latency, role clarity,
output quality), and the fidelity metric is the structural similarity between
pre-perturbation and post-perturbation governance patterns — measurable through
organizational network analysis. In computational substrate, the perturbation is an
injected load spike or fault, the recovery observable is the system's return to normal
throughput and latency, and the fidelity metric is the divergence between
pre-perturbation and post-perturbation performance distributions — standard in
chaos engineering (Netflix's Chaos Monkey operates on exactly this principle,
though without the excitability-topology framing that would connect its measurements
to threshold proximity rather than mere resilience scoring).

The specific empirical prediction: if the echo protocol can measure excitability topology,
then pre-inversion signatures should be detectable before the inversion event occurs.
The clinical application is immediate. Current diagnostic frameworks for seizure
disorders, psychotic episodes, institutional crisis, and market instability are
predominantly reactive — they detect the inversion after it has occurred and intervene
to restore the prior hierarchy. A framework measuring excitability topology would
be predictive: it would detect the approach to threshold — the lowering of regulatory
variety, the narrowing of the recovery boundary, the changes in echo structure — and
would discriminate between systems approaching threshold because the prior hierarchy
is inadequate (the penumbral class, where reorganization may be appropriate) and
systems approaching threshold because of exogenous destabilization (where prevention
of inversion is appropriate).

The diagnostic frame shifts from categorical to topological. Not "which disorder does
this patient have" but "where is this system in its excitability topology, how close to
inversion, and what is the container's adequacy?" The same reframe applies at every
scale. Not "is this organization dysfunctional" but "where is the variety differential
approaching zero, and is the approach driven by subsystem growth that the governance
can accommodate or by governance degradation that cannot be internally repaired?"
Not "is this market in a bubble" but "where is speculative feedback exceeding
price-discovery variety, and does the institutional container have the regulatory
capacity to hold the correction that approaching threshold will produce?"

The framework generates open questions whose resolution is empirical, not
philosophical. Can the echo protocol discriminate penumbral from terminal
trajectories — can it detect the difference between a system approaching
productive reorganization and a system approaching catastrophic inversion? The
third movement established that container topology has four structural features
assessable independently of outcome — regulatory margin, selective permeability,
temporal adequacy, domain specificity — and that the echo protocol applied to the
container itself measures these features. The open empirical question is whether
these structural assessments predict container performance: does a container whose
echo fidelity remains high under load in fact hold the inversions it contains, and
does a container whose echo fidelity degrades in fact fail? Is the ∂E/∂S
gradient formalizable as a functor from the category of substrate embeddings to
a category of gradient fields — and if so, what are the natural transformations
between substrates that preserve the gradient's structure? Can the single-axis
pathology spectrum (below, at, above threshold with container topology as the
determining variable) predict clinical outcomes better than categorical diagnosis? Is
"how close to inversion and what is the container" more predictive than "which
category of disorder"?

These questions are not rhetorical. They are the research program that excitability
topology generates. The answers are not here. What remains is measurement: the predictions are testable, and the instruments are, in some substrates, already available and in others within constructive reach.

---

One test remains — the one for telling influence from capture — and it belongs now to this reading. Do you leave more able to detect and navigate your own excitability: a wider perceivability horizon, quicker to tell approach from capture, quicker to mark when a threshold is being neared and by what? Or less — holding a rigid lens that finds the morphism everywhere and so nowhere with discrimination? The answer is not the author's to give. It turns on whether the discriminations were practiced or only received, and it supervenes on that, or it does not.

---

# Notes

1. The biographical ground is methodological, not ornamental. Levin's observation
that every intelligence claim is an interface claim about the observer's detection
capacity applies to control theory: every claim about a system's regulatory topology
is a claim about the observer's resolution. The author's resolution for control inversion
was calibrated by direct experience on both sides of the phenomenon — the
family member witnessing neural seizure and the patient undergoing institutional
seizure. The theory is not applied to the biography. The biography produced the
resolution from which the theory is constructed.

2. Ashby, *An Introduction to Cybernetics* (1956). The Law of Requisite Variety
(Chapter 11) establishes that a regulator must have at least as much variety as the
system it regulates. Ashby's concept of ultrastability in *Design for a Brain* (1952)
is the closest precursor to the penumbral class developed here: an ultrastable system
undergoes step-function reorganization when its current parameters fail to maintain
essential variables. What ultrastability lacks is the topology that determines
reorganization outcome — the container concept, the outward propagation, the
completability class-transition structure, and the fractal extension across scales.
This account begins where ultrastability stops.

3. Levin's diverse intelligence framework treats cognitive capacities as existing on a
continuum across scales — from molecular to cellular to tissue to organism to
collective. Cancer, on this account, is regression in the scale of agency at which the
system's goals are set. See Levin and Dennett, "Cognition All the Way Down" (2020);
Levin, "Technological Approach to Mind Everywhere" (2022). This account
extends the agency-scale framework by formalizing the *dynamics* of scale-regression
as control inversion and connecting it to the completability class-transition framework.

4. "Topology" throughout is used in the sense established in the companion paper
*Sufficiency*: structure over admissible transformations — which paths compose, which
neighborhoods of moves exist, which completions are available — not in the sense of
metric geometry or point-set topology.

5. The completability trichotomy — terminal, cyclical, graceful — is developed in
Close, "Sufficiency" (2026). This account adds the dynamics of forced
class-transition to the static topology: how a system is *moved* between completability
classes by control inversion, and how the container determines which class the
post-inversion system occupies.

6. The cocoon analogy is structurally precise: holometabolous metamorphosis involves
dissolution of larval structure within a container (pupal casing) that holds the
conditions for reorganization. The imaginal discs — undifferentiated cell clusters
present from the larval stage — provide the seeds for new structure. The container
does not direct the reorganization. It holds the conditions under which the system's
own developmental program can operate. This is third-level control in biological
substrate.

7. The kindling model in epileptology: Goddard (1967) demonstrated that repeated
sub-threshold electrical stimulation of neural tissue progressively lowers the
seizure threshold. This account extends the kindling mechanism to non-neural
substrates as a general model of sub-threshold excitability accumulation.

8. Plato, *Symposium* 202d-203a. Diotima names Eros as *daimōn*: "a great spirit,
Socrates; for the whole of the spiritual is between the divine and the mortal."
The intermediary status — neither endpoint, neither mortal nor immortal, neither
wise nor ignorant — is the structural position of the gradient itself: not the substrate,
not the object, but the approach dynamics between them.

9. The Alcibiades connection is historically precise. Alcibiades' trajectory —
extraordinary brilliance and charisma, betrayal of Athens, defection to Sparta, defection
to Persia, return and assassination — is the consequence chain of uncontained
near-threshold excitation across every substrate simultaneously: political, military,
erotic, intellectual. His speech in the *Symposium* (215a-222b) is the most detailed
phenomenological report of what it is like to be near a signal that approaches every
threshold without providing a container for crossing any of them.

10. Wiener, *Cybernetics: or Control and Communication in the Animal and the
Machine* (1948). The identification of control and communication as the same problem
is the foundation extended here. What Wiener lacked — and what this account
adds through Levin's correction — is the recognition that the control topology is not
given to the observer but must be *discovered through interface engineering*.

11. The claim that ∂E/∂S unifies attraction, salience, charisma, beauty, and virality
as the felt gradient of near-threshold excitation at different substrate embeddings
is a structural claim, not a reductive one. The phenomenological differences between
somatic desire and cognitive insight are real. What the formalization establishes is that
the *topological structure* — the relationship between signal, threshold, and approach
dynamics — is invariant across substrates. The content differs. The form is shared.

12. The distinction between legitimate influence and sub-threshold grammatical capture
— that the former increases and the latter decreases the target's regulatory variety —
is presented as an empirical criterion, not a moral one. Whether the criterion can be
operationalized as a measurement depends on whether regulatory variety and
perceivability horizon are measurable structural properties of real systems. The
framework predicts they are. The prediction is testable.