A Derivation of Intelligence from Guided Relation-Creation

Intelligence is one of those words that gets harder to pin down the more carefully you look at it. Different traditions ground it in different places — computation, language, reasoning, planning, behavior — and the answers don’t converge. Whatever intelligence is, it doesn’t have an obvious natural boundary, and trying to define it directly mostly tells you about whoever’s doing the defining.

So this won’t start with intelligence. It starts somewhere simpler and works up.

The argument exists in two forms. There’s a full mathematical derivation that works through everything with lemmas and proofs. This is the same argument in prose — same moves, less notation, organized around the questions that actually drive it forward.

Why start with viability rather than intelligence?

Because intelligence isn’t the first problem. Life is.

Reality exists. Things exist inside it. But existing doesn’t guarantee continuing to exist. Finite organized systems can fail — they can be broken, starved, poisoned, dissolved, ruptured, disordered, destroyed. A finite organized system continues only when certain conditions hold, and those conditions aren’t optional. They’re what it means for the system to remain itself.

This is more basic than intelligence, and the framework can build upward from it. Whatever intelligence turns out to be, it has to emerge somewhere on the path from finite-organization-that-keeps-going to whatever the upper levels are. We just have to walk it.

What does it mean for a system to continue?

A cell is the clearest starting case. It exists inside reality, fully subject to physics, not exempt from anything. It continues only while certain conditions hold: membrane integrity, usable energy, material exchange, internal chemical order, repair, protection from things that would destroy it.

This is more specific than “the cell exists.” Life is maintained existence — and the maintenance is the cell’s own work. A rock also exists. A rock persists. But a rock’s persistence isn’t performed by the rock; it happens to the rock, as long as the physics happens to hold. The cell is different. The cell does what’s required for its continuation. Its persistence is its own activity.

So the first distinction we need is between persistence that happens to a system and persistence the system performs. Finite organized systems that perform their own persistence are the subjects of what follows.

Call the conditions whose continuation makes the system the kind of thing it is its organization. For a cell, that includes the things named above. For other kinds of systems, it includes their own analogues. We stay general about what specifically counts as a system’s organization, because that varies with the kind of system — but we commit to there being some such structure for any organized system the argument is about.

The system’s viability depends on its relations with reality. Some relations preserve its organization. Others damage it. This distinction is more basic than moral right and wrong, more basic than rational true and false. It’s the primitive difference between what sustains a system and what destroys it. Everything that follows builds on this distinction.

When does motion become action?

A rock can be moved by the world. Wind wears it down. Water carries it. Gravity pulls it from a hill. The rock enters new relations. But the rock doesn’t produce those relations. It’s moved into them.

A living system can produce new relations from itself. A cell opens or closes channels. It absorbs or expels. It repairs. It moves toward or away from things. It divides. Its own organization participates in producing the relation-change.

This is what action is, in the framework’s sense: relation-change generated by the system’s own organization. A stone rolling downhill moves but doesn’t act, because the moving isn’t done by the stone. A cell moving toward nutrient acts, because the moving is what the cell does.

The distinction matters because once a system can act, its activity becomes part of the conditions of its own continuation. Before action, the system is exposed to what reality does to it. After action, it’s also exposed to what it does into reality. It can bring itself into sustaining relations or destructive ones. Toward nutrient or toward poison. Toward repair or toward worse damage.

Why must action be guided?

If the system produces viability-relevant relations through action, and if the same action can sustain in one situation and destroy in another, then action can’t be arbitrary with respect to reality. If nutrient and toxin make no difference to what the system does, it can’t reliably approach one and avoid the other. If tolerable warmth and lethal heat make no difference to what it does, action can’t reliably preserve it.

So action has to be guided. Guidance means viability-relevant differences in reality make a difference to what the system does. The system’s action varies with what matters for its continuation.

This is the first major move. A living system isn’t merely organized. It’s organized in such a way that its own relation-changing activity has to remain sensitive to the reality that sustains or destroys it.

Life is guided relation-creation.

This is a definitional commitment, not a logical consequence. We’re characterizing life by the structural property the argument just showed is necessary for reliable viability. Choosing this definition is what lets everything that follows build.

What changes when guidance becomes mediated by a model?

Guidance can work two ways.

The simple way is immediate. A chemical gradient changes the cell’s internal state; the altered state changes movement. A damaged membrane triggers repair. A concentration difference opens or closes exchange. Reality couples directly to activity, and activity responds. The system doesn’t need to represent the world, think about it, or carry anything across time. It only needs viability-relevant reality to make a difference to what it does.

Immediate guidance is enough for some forms of life. But it’s bounded by what’s happening right now. It can’t explain how a system acts on something absent, remembered, anticipated, or not currently in contact.

For that, guidance has to become mediated. The system carries traces of past contact. It can be altered by past nourishment, past danger, past injury, past success, past failure. It can compare present to retained. Prepare for what hasn’t happened yet. Act on what it’s learned rather than only on what’s pressing on it right now.

When that happens, action isn’t guided just by immediate coupling. It’s guided through a carried structure that lets the system act beyond present contact.

Call that carried structure a model.

A model in this sense is broad. Not necessarily a picture in the head. Not necessarily explicit. Not necessarily symbolic. Not necessarily conscious. A model is any retained, action-guiding structure that lets a system act from more than what’s immediately present. It might be a learned sensitivity, a memory, an expectation, a spatial map, a body schema, a category, a plan, a prediction, a theory, an institutional record.

The threshold for counting as a model: the carried structure has to preserve, organize, anticipate, or generalize viability-relevant reality beyond immediate coupling. A persistent state that just sits there isn’t a model. A structure that lets the system act on the world differently because of what it carries is.

Why does a model create a new kind of danger?

A model is powerful because it frees action from the immediate. The system can act on what’s absent. It can remember. Prepare. Anticipate. Avoid danger before impact. Seek what isn’t currently touching it.

But the same power opens a new structural problem. The model is not reality. Action is selected from the model — from the carried guidance-state — but the action’s actual consequence happens in reality. So the source of action and the field of consequence aren’t identical. The system acts from what it carries. What it does happens in what is real.

When what the system carries loses contact with what’s real, action can stay coherent inside the model while becoming destructive in reality. A stale model can guide action. A narrow model can guide action. A false model can guide action. An overconfident model can guide action. A model protected from correction can guide action. The system acts intelligibly from its representation while its action produces relations the model didn’t anticipate or refuses to incorporate.

This danger is specific to systems that act through models. Immediate guidance doesn’t have it — there’s no gap to grow, because the system is coupled to reality in real time. Mediated guidance has it built in, because the model persists and can be acted from even when reality has changed underneath it.

So model-guided action requires more than a model. It requires contact between model and reality.

What does it mean for the model to remain in contact with reality?

Contact isn’t the model containing all of reality. No finite model can. Contact isn’t certainty. A system can be uncertain and still be in contact. Contact isn’t the model equaling the world — that wouldn’t even be coherent for a finite system inside a larger reality.

Contact is this: viability-relevant differences in reality can still constrain the system’s guidance. When the world differs in ways that matter, those differences can still reach the system’s guidance, the action that follows from it, or the system’s confidence in that action. When that fails, the model is sealed off, and action continues from a representation the world has already falsified.

Put directly: for any difference in reality that matters for action, the system’s guidance has to be able to register a corresponding difference somewhere — in what it represents, in what it does, or in how much it trusts what it does. If it can’t register that difference, the model is acting blindly with respect to that aspect of reality, even when everything seems fine from the outside.

That’s the third major commitment of the argument. Model-guided action remains reality-responsive only while the model stays in contact with reality. The next question is where contact can break.

Where can contact fail?

To find where, we need the minimal structure of model-guided action. Stripped down, it’s a cycle. Reality, the system’s guidance-state, the action guidance selects, the relation that action produces in reality, the system’s later guidance-state — and running alongside the cycle, the system’s estimate of how far its current guidance can be trusted.

Each step in this cycle is a place where viability-relevant reality has to be preserved. Each is a place where preservation can fail.

The first failure-site is entry. Relevant reality has to be able to get into the guidance-state. If a viability-relevant difference exists in the world but doesn’t produce a difference in the model, the system can’t act on it. A toxin is present but isn’t registered. A danger appears but no sign of it enters. A contradiction exists but no evidence of it reaches the model. If entry fails, action starts from a world the system hasn’t received.

The second failure-site is field. Entry isn’t enough. A relevant difference might enter the system and still be placed outside the field the model treats as relevant — counted as background, noise, exception, someone else’s problem. The model represents food and ignores the predator. It represents immediate reward and ignores later exhaustion. It represents production and ignores waste. The system has organized relevance too narrowly. If field fails, the system acts into more reality than it has modeled.

The third failure-site is return. Action changes reality. The relation it produces has to come back to the system as usable guidance. The cell moves but doesn’t register the difference between nourishment and damage. The institution acts but only tracks output, not harm. The model predicts but the result never makes it back. If return fails, action keeps happening while losing contact with what it actually does.

The fourth failure-site is revision. Consequences can return without changing the model. Damage comes back, behavior doesn’t change. Contradiction comes back, the system files it as noise. Correction comes back, gets stored, doesn’t actually update the structure that guides action. Return asks whether the result comes back. Revision asks whether what comes back can change future guidance. If revision fails, the system is informed without being corrected.

The fifth failure-site is measure. Even a model whose entry, field, return, and revision are working can mismeasure its own incompleteness. Weak evidence, high confidence. Strong evidence, treated as unusable. Knowing something locally and applying it globally. The problem isn’t just that the model is incomplete — it’s that the system mismeasures its incompleteness. If measure fails, action is no longer proportioned to the contact the system has actually earned.

So: five sites. Entry, field, return, revision, measure.

Why does this structure close at five and not more?

The five weren’t chosen by preference. They came out of the cycle.

Reality has to affect guidance — that’s entry. Guidance has to include the field its action enters — that’s field. The relation produced by action has to be available to future guidance — that’s return. Returned relation has to be able to alter the guidance-state — that’s revision. The system has to estimate the strength and limit of its guidance — that’s measure.

So a proposed sixth contact-site has only two possible forms. Either the minimal cycle is missing a necessary term, or one of the existing terms contains another irreducible contact-relation hiding inside it. If neither is true, the proposal isn’t a sixth contact-site.

Lots of things matter for model-guided systems besides these five.

Memory matters for any guidance to persist across time — but memory isn’t a contact-relation. It’s what the existing contact-relations require to operate beyond an instant.

Execution fidelity matters for whether the chosen action actually gets performed — but execution is implementation, not contact.

Goal-formation matters for what action gets chosen — but goals direct rather than constrain reality-contact.

Attention matters for what gets brought into the model — but attention is a selection mechanism within entry and field.

Power matters because it scales effect — but power multiplies contact or its absence rather than constituting it.

Each of these belongs to a neighboring structure. Memory to persistence. Execution to implementation. Goals to direction. Attention to selection. Power to capability. None of them adds a new place where viability-relevant reality has to be preserved if action is to stay reality-guided. They surround the contact-structure. They don’t expand it.

That’s what closure means here. The five sites aren’t a complete account of everything that matters for an intelligent system. They’re the complete account of where model-world contact can fail. The structure is closed against expansion of that specific question. Other questions are real but need their own structures.

What does intelligence become, given all of this?

Walking back through the path: finite organization that performs its own maintenance; action as relation-change the system produces from itself; guidance as the requirement that action varies with reality; immediate guidance for some forms of life; mediated guidance through carried structures called models; the model-gap as the new danger that mediation creates; contact as what closes that gap; and the five sites where contact has to be maintained.

Intelligence is now sayable in the framework’s terms:

Intelligence is contact-closed model-guided relation-creation.

Plainer: intelligence is the capacity of a finite system to guide its relation-changing action through a model while keeping that model answerable to the reality where the action lands.

This doesn’t require carbon. Or neurons. Or language. Or consciousness as a starting assumption. It requires the structural features — model-mediated guidance and the maintenance of contact across the five sites. So it applies to a cell whose chemistry implements all this, to a person whose nervous system implements it, to an institution whose procedures implement it, to an artificial system whose architecture implements it. Wherever the structure obtains.

What this definition makes visible is what intelligence isn’t.

Intelligence isn’t the production of effects.

A system can be enormously capable — manipulating symbols, optimizing toward targets, generating output, acting quickly and forcefully — and still fail to be intelligent in this sense. If reality can’t enter its guidance, or if it can’t include the field its action affects, or if consequences don’t return, or if returned reality can’t revise it, or if it can’t measure the limits of its own contact, then it’s acting in the world without keeping its action answerable to what the world actually is. That’s capability without closed contact.

The distinction is sharp once it’s stated. Capability is the power to produce effects. Intelligence is the power to produce effects while remaining in contact with what those effects do. A capable system without contact produces effects rapidly and confidently — blindly with respect to the field its action enters. A capable system with contact produces effects whose consequences inform what it produces next. The first is power without intelligence. The second is power that intelligence is keeping faithful.

This matters in practice. Many of the systems with the most effect in the world right now — markets, platforms, large institutions, scaled algorithms — are extremely capable. Whether they’re intelligent in this sense is a separate question. The question is whether their actions stay in contact with what their actions actually do, across all five sites. Whether the relevant reality can enter their guidance. Whether their representation of what they affect includes the actually affected field. Whether consequences return. Whether returned consequence can change their structure. Whether they accurately measure how much of their contact they’ve actually earned.

The framework doesn’t answer that for any specific system. But it tells you what to look for. And where contact most often breaks. And why a system losing contact at any of these sites becomes more powerful and more blind at the same time.

That’s what intelligence is. And that’s what its absence looks like when capability continues without it.

---

1. Starting Point: Finite Organization and Continuation

We do not begin with intelligence.

We begin with something lower: a finite organized system inside reality.

A finite organized system does not continue merely because it exists. It can remain organized, or it can fail. It can be broken, starved, poisoned, dissolved, disordered, ruptured, or destroyed.

So the first problem is not intelligence.

The first problem is continuation.

$\boxed{\text{How can a finite organized system continue within a reality that does not guarantee its continuation?}}$

Let:

$S$

be a finite organized system.

Let:

$W$

be the relevant world-condition in which $S$ exists.

The system is finite because:

$S \neq W.$

That is, $S$ is not identical with the whole world-condition in which it exists.

The system is organized because there is some structure whose preservation distinguishes $S$ continuing as itself from $S$ failing, dissolving, or becoming something else.

Call that structure:

$O(S).$

So:

$\boxed{O(S)=\text{the organization whose preservation is required for }S\text{ to continue as }S.}$

Thus:

$\boxed{\operatorname{FiniteOrg}(S)\Rightarrow \exists O(S).}$

This is the starting domain of the derivation. It does not yet assume action, guidance, model, contact, or intelligence. It only assumes finite organization under possible continuation or failure.

---

2. Viability

Given a finite organized system $S$, we can define viability.

Viability is not an added moral value. It is the condition under which the system’s organization continues.

Let:

$V(S,W)=1$

mean that $S$ remains viable in world-condition $W$.

That is:

$\boxed{V(S,W)=1 \iff O(S)\text{ is preserved under }W.}$

Let:

$V(S,W)=0$

mean that $S$ is not viable in world-condition $W$.

That is:

$\boxed{V(S,W)=0 \iff O(S)\text{ is damaged, destabilized, or ended under }W.}$

So:

$\boxed{\text{viability}=\text{the preservation of the organizing conditions by which }S\text{ continues as }S.}$

A finite organized system may be viable under one world-condition and non-viable under another:

$V(S,W_1)=1$

while:

$V(S,W_2)=0.$

This means reality makes a difference to whether the system continues.

At this stage, $V(S,W)\in\{0,1\}$ is only a minimal contrast between preservation and failure. It is not yet a numerical measure of welfare, value, health, or flourishing.

---

3. Relations

A finite organized system does not exist outside reality. It exists in relation to the world-condition in which its continuation or failure occurs.

Let:

$Rel(S,W)$

denote the relation between $S$ and $W$.

Some relations preserve or restore the organization of $S$. Other relations damage or destroy it.

Define:

$\operatorname{SupportsV}(Rel(S,W))$

when:

$Rel(S,W)$

contributes to:

$V(S,W)=1.$

Define:

$\operatorname{DamagesV}(Rel(S,W))$

when:

$Rel(S,W)$

tends toward:

$V(S,W)=0.$

So we get the first continuation-relative distinction:

$\boxed{\text{some relations are viability-supporting, and some relations are viability-damaging.}}$

This is not yet moral good and bad.

It is not yet rational true and false.

It is only the structural distinction between relations that preserve the system’s organization and relations that damage or destroy it.

Plainly:

$\boxed{\text{at this level, “good-for” means good-for-continuation.}}$

---

4. Relation-Fit Lemma

If one world-condition preserves the system and another world-condition destroys or destabilizes it, then the system’s continuation cannot be explained by the system alone.

It depends on relation-fit between the system and reality.

$\boxed{\textbf{Relation-Fit Lemma}}$

Let $S$ be a finite organized system with organization $O(S)$.

Let $W_1$ and $W_2$ be world-conditions.

Assume:

$V(S,W_1)=1$

and:

$V(S,W_2)=0.$

Then the continuation of $S$ depends on relation-fit between $S$ and the relevant world-condition.

Formally:

$\boxed{V(S,W_1)=1 \land V(S,W_2)=0 \Rightarrow \operatorname{RelFitNeed}(S).}$

where:

$\boxed{\operatorname{RelFitNeed}(S) = \text{the continuation of }S\text{ requires relation-fit with reality.}}$

Proof

Since:

$V(S,W_1)=1,$

the organization $O(S)$ is preserved under $W_1$.

So:

$\operatorname{SupportsV}(Rel(S,W_1)).$

Since:

$V(S,W_2)=0,$

the organization $O(S)$ is not preserved under $W_2$.

So:

$\operatorname{DamagesV}(Rel(S,W_2)).$

Therefore the continuation of $S$ varies with how $S$ stands in relation to the relevant world-condition.

Thus:

$\boxed{\operatorname{RelFitNeed}(S).}$

Plainly:

$\boxed{\text{where continuation varies by condition, continuation depends on relation-fit.}}$

---

5. Action

So far we have shown that a finite organized system depends on relation-fit with reality.

But relation-fit alone is not yet action.

A rock also exists in relation to reality. Wind can wear it down. Water can carry it. Gravity can pull it from a hill. The rock enters new relations, but those relation-changes are imposed on it.

Some systems are different. Their own organization can participate in producing relation-change.

A cell can open or close channels. It can absorb or expel. It can repair. It can move toward or away. It can alter its boundary-relation to the environment.

This gives the action gate.

$\boxed{\text{Action is relation-change generated through the system’s own organization.}}$

Let:

$a$

be an action available to $S$.

Let:

$R(S,W,a)$

be the relation produced when $S$ performs action $a$ in world-condition $W$.

So:

$\boxed{R(S,W,a)=\text{the relation between }S\text{ and }W\text{ produced by action }a.}$

Define:

$\boxed{\operatorname{Action}(a,S) \iff a\text{ is relation-change generated through }S\text{'s own organization.}}$

This definition does not require consciousness, intention, language, or deliberation.

It requires only that the system’s own organization participates in producing the relation-change.

So action is not mere motion.

$\boxed{\text{Motion is relation-change. Action is system-generated relation-change.}}$

---

6. The Guidance Problem

Once a system can act, its own activity becomes part of the conditions of its continuation.

Before action, the system is exposed to what reality does to it.

After action, the system is also exposed to what it does into reality.

It can bring itself into sustaining relations.

It can also bring itself into damaging relations.

The same action may preserve the system in one world-condition and damage it in another. Moving toward a chemical source is viable if the source is nutrient. It is damaging if the source is poison. Opening a boundary is viable if what enters restores the system. It is damaging if what enters destroys it.

So action creates a new problem:

$\boxed{\text{How can self-produced relation-change remain compatible with continued existence?}}$

If the viability-status of action depends on the world-condition, then action cannot be arbitrary with respect to reality.

If nutrient and toxin make no difference to what the system does, the system cannot reliably approach one and avoid the other.

If tolerable warmth and lethal heat make no difference to what the system does, the system cannot reliably preserve itself across those conditions.

So define guidance:

$\boxed{\operatorname{Guided}(S)}$

to mean that viability-relevant differences in reality can make action-relevant differences in what $S$ does.

Let:

$\Delta W_V$

mean a viability-relevant difference in world-condition.

Let:

$\Delta a$

mean an action-relevant difference in what the system does.

Then guidance has the form:

$\boxed{\Delta W_V \rightsquigarrow \Delta a.}$

The symbol:

$\rightsquigarrow$

means “can constrain, guide, or make a relevant difference to.” It does not mean strict deterministic one-to-one causation.

Plainly:

$\boxed{\text{Guidance means action remains sensitive to the differences in reality that matter for continuation.}}$

---

7. Guidance-Need Lemma

If an acting system faces two world-conditions where the same action is viable in one and damaging in another, then reliable viability across those conditions requires action-relevant differentiation between them.

$\boxed{\textbf{Guidance-Need Lemma}}$

Let $S$ be a finite organized system.

Let $a_1$ be an action available to $S$.

Let $W_1$ and $W_2$ be world-conditions.

Assume that action $a_1$ produces a viable relation in $W_1$:

$Viable(R(S,W_1,a_1)).$

Assume that the same action $a_1$ does not produce a viable relation in $W_2$:

$\neg Viable(R(S,W_2,a_1)).$

Then reliable viability across $W_1$ and $W_2$ requires the difference between $W_1$ and $W_2$ to be action-relevant for $S$.

Formally:

$\boxed{Viable(R(S,W_1,a_1)) \land \neg Viable(R(S,W_2,a_1)) \Rightarrow Need_S(\Delta W_V \rightsquigarrow \Delta a).}$

where:

$\boxed{Need_S(\Delta W_V \rightsquigarrow \Delta a) = S\text{ requires viability-relevant world-differences to be able to guide action for reliable viability.}}$

Proof

By assumption:

$Viable(R(S,W_1,a_1)).$

So action $a_1$ is viability-supporting in $W_1$.

By assumption:

$\neg Viable(R(S,W_2,a_1)).$

So action $a_1$ is not viability-supporting in $W_2$.

Therefore the viability-status of the same action differs across $W_1$ and $W_2$.

Suppose the difference between $W_1$ and $W_2$ cannot make any action-relevant difference for $S$:

$\neg(\Delta W_V \rightsquigarrow \Delta a).$

Then $S$ has no action-guiding basis, from that difference, for treating $W_1$ and $W_2$ differently.

So $S$ may produce the same action $a_1$ in both conditions.

But $a_1$ is viable in $W_1$ and non-viable in $W_2$.

Therefore $S$ cannot reliably preserve viability across $W_1$ and $W_2$ unless the difference between those conditions can guide action.

Thus:

$\boxed{Need_S(\Delta W_V \rightsquigarrow \Delta a).}$

Plainly:

$\boxed{\text{reliable action-based viability requires guidance.}}$

Guidance is not yet thought, representation, or intelligence. It is the structure by which viability-relevant reality-differences can make action-relevant differences.

---

Interim Result

We have established:

$\boxed{\operatorname{FiniteOrg}(S) \Rightarrow \exists O(S)}$

$\boxed{O(S)+\text{possible preservation/failure} \Rightarrow V(S,W)}$

$\boxed{V(S,W_1)=1 \land V(S,W_2)=0 \Rightarrow \operatorname{RelFitNeed}(S)}$

$\boxed{\operatorname{Action}(a,S) \Rightarrow R(S,W,a)}$

and:

$\boxed{Viable(R(S,W_1,a_1)) \land \neg Viable(R(S,W_2,a_1)) \Rightarrow Need_S(\Delta W_V \rightsquigarrow \Delta a).}$

So the derivation has reached:

$\boxed{\text{finite organization} \Rightarrow \text{viability} \Rightarrow \text{relation-fit} \Rightarrow \text{action} \Rightarrow \text{guidance need.}}$

A model has not yet been introduced.

8. Immediate Guidance

Guidance does not yet require a model.

A system can be guided immediately when a difference in reality directly alters the system’s state, and that altered state changes action.

A chemical gradient may alter a cell’s internal state, and that altered state may change movement. A damaged boundary may trigger repair. A concentration difference may open or close exchange.

In this case, the system does not need to represent the world as an object. It does not need to think, judge, believe, predict, or plan.

It only needs viability-relevant reality to make a difference to what it does.

Let:

$\Delta W_V$

mean a viability-relevant difference in world-condition.

Let:

$\Delta S$

mean a relevant difference in the system’s internal state.

Let:

$\Delta a$

mean a relevant difference in action.

Immediate guidance has the form:

$\boxed{\Delta W_V \rightsquigarrow \Delta S \rightsquigarrow \Delta a.}$

This means:

$\boxed{\text{a viability-relevant difference in reality can alter the system, and that alteration can alter action.}}$

So immediate guidance is:

$\boxed{\operatorname{ImmediateGuidance}(S)}$

when:

$\boxed{\Delta W_V \rightsquigarrow \Delta S \rightsquigarrow \Delta a.}$

Plainly:

$\boxed{\text{Immediate guidance is direct reality-action coupling through the system’s present state.}}$

Immediate guidance may be enough for some forms of life. But it does not yet explain how a system can act from what is absent, remembered, expected, hidden, delayed, possible, or abstract.

For that, guidance must become mediated.

---

9. Mediated Guidance

Immediate guidance is limited by present coupling.

A system may also carry traces of reality across time.

It may retain the result of prior contact. It may be altered by past nourishment, past danger, past injury, past success, or past failure. It may compare present conditions to retained traces. It may prepare for what is not yet present. It may act from what has been learned, not only from what is immediately pressing on it.

At that point, action is no longer guided only by immediate coupling.

It is guided through something the system carries.

Let:

$G_t$

be a guidance-state of system $S$ at time $t$.

A guidance-state becomes mediated when it can guide action beyond immediate coupling alone.

So mediated guidance has the form:

$\boxed{G_t \rightsquigarrow a_t.}$

Or, when present reality also contributes:

$\boxed{(W_t,G_t)\rightsquigarrow a_t.}$

This means the action may be shaped both by current world-condition and by a carried guidance-state.

Mediated guidance is therefore not the rejection of immediate guidance. It is the addition of a carried action-guiding structure.

Define:

$\boxed{\operatorname{MediatedGuidance}(S)}$

when $S$‘s action is selected through a carried guidance-state $G_t$ that goes beyond immediate coupling alone.

Formally:

$\boxed{\operatorname{MediatedGuidance}(S) \iff \exists G_t\,[G_t\text{ is carried by }S \land G_t \rightsquigarrow a_t \text{ beyond immediate coupling alone}].}$

Plainly:

$\boxed{\text{Mediated guidance begins when the system acts through a carried structure.}}$

The carried structure need not be conscious, linguistic, symbolic, or explicit. What matters is its role in action-selection.

---

10. Model

We can now define model.

A model is not necessarily a picture in the head.

It is not necessarily a theory.

It is not necessarily linguistic, symbolic, conscious, or explicit.

A model is any carried guidance-structure through which action is selected beyond immediate coupling alone.

So:

$\boxed{\text{A model is a carried guidance-structure through which action is selected beyond immediate coupling alone.}}$

Let:

$\operatorname{Model}(G_t,S)$

mean that $G_t$ functions as a model for system $S$.

Then:

$\boxed{\operatorname{Model}(G_t,S) \iff G_t\text{ is carried by }S \land G_t\text{ guides action} \land G_t\text{ operates beyond immediate coupling alone.}}$

This includes simple and complex cases.

A model may be a learned sensitivity, a memory, an expectation, a map, a body schema, a category, a plan, a prediction, a theory, or an institutional record.

The important feature is not the material form of the model.

The important feature is its role:

$\boxed{G_t \rightsquigarrow a_t.}$

A physical trace counts as model-level only when it mediates action by preserving, organizing, anticipating, or generalizing action-relevant reality beyond immediate coupling.

So:

$\boxed{\text{model} \neq \text{any physical trace.}}$

A model is an action-guiding carried structure.

Plainly:

$\boxed{\text{A model lets the system act from what it carries, not only from what is immediately present.}}$

But a model can guide action while being stale, narrow, false, incomplete, or overconfident.

That creates the next problem.

---

11. The Model-Gap

A model is powerful because it frees action from immediate coupling.

The system can act on what is absent, remembered, expected, possible, hidden, delayed, or generalized.

But this power creates a structural danger.

The model is not reality.

$\boxed{G_t \neq W_t.}$

The action is selected from guidance:

$\boxed{a_t=A(G_t).}$

More generally, where immediate entry and model-guidance both contribute:

$\boxed{a_t=A(W_t,G_t).}$

But the action’s consequence occurs in reality:

$\boxed{R_t=R(S,W_t,a_t).}$

Here:

$R_t$

is the relation produced when $S$ acts in world-condition $W_t$ by action $a_t$.

So we have the structural gap:

$\boxed{\text{selection occurs from }G_t,}$

while:

$\boxed{\text{consequence occurs in }W_t.}$

The system acts from what it carries.

But what it does happens in what is real.

This is the model-gap:

$\boxed{\operatorname{ModelGap}(S,t)}$

where:

$\boxed{\operatorname{ModelGap}(S,t) \iff G_t\neq W_t \land a_t=A(G_t)\text{ or }A(W_t,G_t) \land R_t=R(S,W_t,a_t).}$

Plainly:

$\boxed{\text{Model-guided action is selected from guidance, but tested by reality.}}$

This is why model-guided action requires more than a model.

A stale model can guide action.

A narrow model can guide action.

A false model can guide action.

An overconfident model can guide action.

A model protected from correction can guide action.

The model may remain coherent inside itself while the action it selects becomes damaging in reality.

So model-guided action requires contact between model and reality.

Before deriving contact, one further distinction must be made explicit.

---

12. Field-Reality Priority

The model-gap already tells us that action is selected from guidance, while consequence occurs in reality.

This means return is not what makes a consequence real.

A consequence may occur in the field whether or not it later returns to the actor’s guidance-state.

Let:

$R_t=R(S,W_t,a_t)$

be the produced relation in reality.

Let:

$G_{t+1}$

be the later guidance-state of the system.

Return, when it occurs, has the form:

$R_t \rightsquigarrow G_{t+1}.$

But the occurrence of $R_t$ does not depend on that return.

So:

$\boxed{R_t \text{ occurs in }W_t}$

does not require:

$\boxed{R_t \rightsquigarrow G_{t+1}.}$

Therefore:

$\boxed{R_t \nrightsquigarrow G_{t+1} \not\Rightarrow \neg R_t.}$

Plainly:

$\boxed{\text{unreturned consequence is still field-real consequence.}}$

Or:

$\boxed{\text{return reconnects guidance to consequence; it does not create consequence.}}$

This is the Field-Reality Priority Lemma.

$\boxed{\textbf{Field-Reality Priority Lemma}}$

If action $a_t$ produces a relation $R_t$ in world-condition $W_t$, then $R_t$‘s occurrence does not depend on $R_t$‘s return to the actor’s later guidance-state.

Formally:

$\boxed{R_t=R(S,W_t,a_t) \Rightarrow \left[ R_t \nrightsquigarrow G_{t+1} \not\Rightarrow \neg R_t \right].}$

Proof

By the model-gap, action is selected from guidance:

$a_t=A(G_t)$

or more generally:

$a_t=A(W_t,G_t).$

The produced relation is:

$R_t=R(S,W_t,a_t).$

Therefore $R_t$ is a relation produced in the world-condition $W_t$.

Return is a further relation from the produced consequence to later guidance:

$R_t \rightsquigarrow G_{t+1}.$

A further relation may fail without negating the prior occurrence of $R_t$.

So:

$\boxed{R_t \nrightsquigarrow G_{t+1} \not\Rightarrow \neg R_t.}$

Plainly:

$\boxed{\text{a consequence can be real in the field without returning to the actor's guidance.}}$

The world does not wait for the actor to notice what the actor’s action has done.

If return fails, the consequence has not failed to exist.

The actor’s contact has failed.

---

Interim Result

We have now reached:

$\boxed{\text{guidance} \Rightarrow \text{immediate or mediated guidance} \Rightarrow \text{model} \Rightarrow \text{model-gap} \Rightarrow \text{field-reality priority.}}$

More explicitly:

$\boxed{\Delta W_V \rightsquigarrow \Delta S \rightsquigarrow \Delta a}$

gives immediate guidance.

$\boxed{G_t \rightsquigarrow a_t}$

gives mediated guidance.

$\boxed{\operatorname{Model}(G_t,S)}$

holds when $G_t$ is a carried guidance-structure through which action is selected beyond immediate coupling.

Since:

$\boxed{G_t\neq W_t}$

and:

$\boxed{a_t=A(G_t)}$

while:

$\boxed{R_t=R(S,W_t,a_t),}$

we get:

$\boxed{\operatorname{ModelGap}(S,t).}$

And since:

$\boxed{R_t \nrightsquigarrow G_{t+1} \not\Rightarrow \neg R_t,}$

we get:

$\boxed{\text{unreturned consequence is still field-real consequence.}}$

The next task is to derive contact: the structure by which model-guided action remains answerable to the reality where its consequences occur.

13. Contact

Model-guided action contains a structural gap.

The system acts from a guidance-state:

$G_t$

but the produced relation occurs in reality:

$R_t=R(S,W_t,a_t).$

Because:

$G_t \neq W_t.$

the model cannot simply be assumed to match reality.

So model-guided action requires a way for relevant reality to continue constraining guidance.

That condition is contact.

$\boxed{\text{Contact is the preservation of action-relevant, viability-relevant difference between reality and guidance.}}$

Contact does not mean that the model contains all of reality. No finite model can.

Contact does not mean certainty. A system can be uncertain and still remain in contact.

Contact does not mean exact equality between guidance and world:

$G_t = W_t.$

Contact means that differences in reality that matter for viable action can still make a difference to guidance, action, or the system’s estimate of how far its guidance can be trusted.

Let:

$\Delta W_V$

mean a viability-relevant difference in world-condition.

Let:

$\Delta G$

mean a relevant difference in guidance.

At the simplest level, contact has the form:

$\boxed{\Delta W_V \rightsquigarrow \Delta G.}$

But model-guided action involves more than the guidance-state alone. Relevant reality may need to alter:

$G$

the guidance-state,

$a$

the action selected,

or:

$K$

the system’s estimate of the strength, reach, or limit of its guidance.

So more generally:

$\boxed{\Delta W_V \rightsquigarrow \Delta(G,a,K).}$

Here:

$K$

does not yet mean a numerical measure.

It means:

$\boxed{K=\text{the system's estimate, however primitive, of the reliability, scope, or limit of its guidance.}}$

So define:

$\boxed{Contact(G,W,a)}$

as the condition under which viability-relevant differences in $W$, relative to action $a$, can still constrain the system’s guidance, action, or confidence in guidance.

Formally:

$\boxed{Contact(G,W,a) \iff \Delta W_V \rightsquigarrow \Delta(G,a,K)}$

where the relevant $\Delta W_V$ are those differences in reality that matter for the viable guidance of action $a$.

Plainly:

$\boxed{\text{A model remains in contact when reality can still correct, constrain, or calibrate the action it guides.}}$

---

14. Contact-Need Lemma

If a system selects action through guidance, and if different world-conditions require different actions for viability, then the action-selecting guidance process must preserve the relevant difference between those world-conditions.

Otherwise the system may act the same way where reality requires different action.

To avoid over-narrowing the proof, we do not assume that action is selected only from $G$.

Earlier, we allowed:

$a_t=A(G_t)$

and also the more general case:

$a_t=A(W_t,G_t).$

So define:

$H(W)$

as the effective action-selecting guidance condition of $S$ under world-condition $W$.

This includes whatever actually participates in selecting action: carried guidance, model-state, immediate entry, current coupling, or any relevant combination of these.

So:

$\boxed{H(W)=\text{the effective guidance condition from which action is selected under }W.}$

Then action-selection has the form:

$\boxed{a=A(H(W)).}$

Now we can state the lemma.

$\boxed{\textbf{Contact-Need Lemma}}$

Let $S$ be a model-guided system.

Let $W_1$ and $W_2$ be world-conditions.

Let $H(W_1)$ and $H(W_2)$ be the effective guidance conditions of $S$ under those world-conditions.

Assume there is an action $a_1$ such that:

$Viable(R(S,W_1,a_1))$

and:

$\neg Viable(R(S,W_2,a_1)).$

Then reliable model-guided viability across $W_1$ and $W_2$ requires the effective guidance condition to preserve the relevant difference between $W_1$ and $W_2$.

Formally:

$\boxed{Viable(R(S,W_1,a_1)) \land \neg Viable(R(S,W_2,a_1)) \Rightarrow Need_S(H(W_1)\neq H(W_2)).}$

where:

$\boxed{Need_S(H(W_1)\neq H(W_2)) = S\text{ requires an action-relevant difference in guidance corresponding to the viability-relevant difference in reality.}}$

Proof

Since $S$ is model-guided, action is selected through an effective guidance condition:

$a=A(H(W)).$

By assumption:

$Viable(R(S,W_1,a_1)).$

So $a_1$ is viable in $W_1$.

By assumption:

$\neg Viable(R(S,W_2,a_1)).$

So $a_1$ is not viable in $W_2$.

Suppose the effective guidance condition does not preserve the relevant difference between $W_1$ and $W_2$:

$H(W_1)=H(W_2).$

Since action is selected from $H(W)$, equal effective guidance conditions select the same action:

$H(W_1)=H(W_2) \Rightarrow A(H(W_1))=A(H(W_2)).$

So the system has no action-selecting basis, within guidance, for selecting differently across $W_1$ and $W_2$.

But by assumption, the same action is not viable in both conditions.

Therefore reliable viability across $W_1$ and $W_2$ requires the effective guidance condition to preserve the viability-relevant difference between them:

$\boxed{Need_S(H(W_1)\neq H(W_2)).}$

Plainly:

$\boxed{\text{If two conditions require different action, guidance must not collapse them into the same action-selecting state.}}$

Where reality makes a viability-relevant difference, guidance must preserve that difference in an action-relevant way.

So:

$\boxed{\text{model-guided viability requires contact with viability-relevant reality.}}$

---

15. Contact-Failure

Contact is the preservation of action-relevant, viability-relevant difference between reality and guidance.

So contact fails when such difference is lost, excluded, blocked, distorted, ignored, or mismeasured.

Let:

$\Delta W_V$

mean a viability-relevant difference in reality.

For model-guided action to remain reality-guided, that difference must be able to affect the guidance process where it matters.

In the simplest case:

$\Delta W_V \rightsquigarrow \Delta G.$

But model-guided action is not only a relation between world and guidance-state.

The system receives reality, selects action, produces a relation in the world, receives or fails to receive consequences, revises or fails to revise, and estimates how far its guidance can be trusted.

So the more general contact condition is:

$\boxed{\Delta W_V \rightsquigarrow \Delta(G,a,R,G',K).}$

where:

$G=\text{guidance-state}$ $a=\text{action selected}$ $R=\text{relation produced in reality}$ $G'=\text{later guidance-state}$ K=, reach, and limit.K=\text{, reach, and limit.}K=estimate of guidance-strength, reach, and limit.

Thus:

$\boxed{\text{Contact-failure occurs when a reality-difference relevant to viable action fails to constrain the guidance process where it must.}}$

Formally:

$\boxed{ContactFailure(G,W,a) \iff \exists \Delta W_V [ \Delta W_V \text{ is action-relevant} \land \Delta W_V \nrightsquigarrow \Delta(G,a,R,G',K) ].}$

Plainly:

$\boxed{\text{contact fails when reality matters, but the guidance process does not let it matter.}}$

---

16. Minimal Structure of Model-Guided Action

To identify where contact can fail, we need the minimal structure of model-guided action.

There is a world-condition:

$W.$

There is a guidance-state:

$G.$

There may also be immediate entry from the current world-condition.

So, more generally, define:

$H(W)$

as the effective action-selecting guidance condition under $W$.

This includes whatever actually participates in action-selection: carried model-state, immediate input, current coupling, or their combination.

Action is selected from that effective guidance condition:

$\boxed{a=A(H(W)).}$

The action produces a relation in reality:

$\boxed{R=R(S,W,a).}$

There is then a later guidance-state:

G′G’G′

or, more broadly, a later effective guidance condition:

H′.H’.H′.

And because finite guidance is never total, the system has some estimate of how far its guidance can be trusted:

$K(H,a).$

So the minimal contact-structure is:

$\boxed{W \rightsquigarrow H \rightarrow a \rightarrow R \rightsquigarrow H'}$

with:

$\boxed{K(H,a).}$

For readability, we may still write:

$W \rightarrow G \rightarrow a \rightarrow R \rightarrow G'$

when $G$ is functioning as the action-guiding condition.

But the hardened form is:

$\boxed{W \rightsquigarrow H \rightarrow a \rightarrow R \rightsquigarrow H'.}$

Plainly:

$\boxed{\text{Reality must enter guidance; guidance selects action; action produces a real relation; the result may return; guidance may revise; confidence must be calibrated.}}$

The contact-sites are the places where viability-relevant reality must be preserved in this structure.

17. Entry

The first contact-site is entry.

Entry is the condition that relevant reality can enter the guidance process.

If a viability-relevant difference exists in the world, but produces no relevant difference in guidance, then the system cannot act according to that difference.

Formally:

$\boxed{Entry(G,W)}$

means:

$\boxed{\Delta W_V \rightsquigarrow \Delta H}$

or, in model-level terms:

$\boxed{\Delta W_V \rightsquigarrow \Delta G.}$

Entry failure occurs when:

$\boxed{\Delta W_V \nrightsquigarrow \Delta H.}$

Plainly:

$\boxed{\text{entry fails when relevant reality cannot get into the guidance process.}}$

A toxin is present, but the system does not register it.

A danger appears, but no usable sign of it enters.

A contradiction exists, but no evidence of it reaches the model.

An injury occurs, but no signal of damage reaches guidance.

In each case, the world differs in a way that matters, but guidance does not differ with it.

So the first contact-condition is:

$\boxed{\text{relevant reality must be able to enter guidance.}}$

Entry is necessary, but not sufficient. A difference may enter the system and still be ignored, mis-scoped, blocked from return, prevented from revising guidance, or mismeasured.

---

18. Field

Entry is not enough.

A relevant difference may enter the system but be placed outside the field the guidance-state treats as relevant.

This is field failure.

It is not simple blindness. It is mis-scoping.

The system may receive something, but treat it as background, noise, exception, externality, or outside the action’s concern.

Let:

$D(a,W)$

be the real field affected by action $a$ under world-condition $W$.

Let:

$F_G(a)$

be the field treated as relevant by the guidance-state.

Then field failure occurs when:

$\boxed{F_G(a)\subset D(a,W)}$

and the omitted part contains action-relevant, viability-relevant differences.

More explicitly:

$\boxed{FieldFailure(G,W,a) \iff \exists x[ x\in D(a,W) \land x\notin F_G(a) \land x\text{ carries action-relevant viability-difference} ].}$

Plainly:

$\boxed{\text{field fails when the system acts into more reality than its guidance includes.}}$

The model represents food but omits the predator near it.

The model represents immediate reward but omits later exhaustion.

The model represents production but omits waste.

The model represents the intended target of action but omits the wider field disturbed by action.

In each case, the problem is not that nothing entered. The problem is that relevance was organized too narrowly.

So the second contact-condition is:

$\boxed{\text{guidance must include the relevant field affected by action.}}$

Field does not mean the model must include all reality. Field is action-relative.

The relevant field is:

$D(a,W),$

not the whole of reality.

So:

$\boxed{\text{field-contact requires inclusion of the affected field, not omniscience.}}$

---

19. Return

Action produces a relation in reality.

Let:

$R_t=R(S,W_t,a_t)$

be the relation produced by action $a_t$ in world-condition $W_t$.

By Field-Reality Priority:

$\boxed{R_t\text{ is real in }W_t\text{ whether or not it returns to guidance.}}$

Return does not create the consequence.

Return reconnects guidance to a consequence already real in the field.

So return is the condition that the produced relation can become available to later guidance.

Formally:

$\boxed{Return(R_t,G_{t+1})}$

means:

$\boxed{R_t \rightsquigarrow G_{t+1}.}$

More generally:

$\boxed{R_t \rightsquigarrow H_{t+1}.}$

Return failure occurs when:

$\boxed{R_t \nrightsquigarrow G_{t+1}.}$

But:

$\boxed{R_t \nrightsquigarrow G_{t+1} \not\Rightarrow \neg R_t.}$

Plainly:

$\boxed{\text{return fails when the consequence exists, but cannot get back into guidance.}}$

The cell moves, but receives no usable difference between nourishment and damage.

The animal acts, but cannot register success, injury, or changed condition.

The institution acts, but tracks output and not harm.

The model predicts, but the result never reaches the model.

In each case, action has consequences. But those consequences vanish from guidance.

So the third contact-condition is:

$\boxed{\text{the consequences of action must be able to return to guidance.}}$

Return is required for actor-contact, not for the consequence to be real.

---

20. Revision

Return is still not enough.

A consequence may return, but the guidance-state may not be changeable by it.

Reality comes back, but the model does not move.

This is revision failure.

Let:

$R_t \rightsquigarrow G_{t+1}$

mean the consequence returns to guidance.

Revision means that returned reality can alter guidance when alteration is required.

Let:

$Rev(G_t,R_t)=G_{t+1}$

mean that guidance is revised in light of returned consequence $R_t$.

Revision succeeds when returned reality can change future guidance where change is required.

Revision failure occurs when:

$\boxed{R_t \rightsquigarrow G_{t+1}}$

but:

$\boxed{G_{t+1}=G_t}$

even though $R_t$ shows that $G_t$ is wrong, incomplete, unsafe, or misweighted.

Plainly:

$\boxed{\text{revision fails when reality returns but guidance cannot change with it.}}$

The system receives damage but does not alter behavior.

It receives failure but protects the old pattern.

It receives contradiction but treats it as noise.

It receives correction but stores it without changing the structure that guides future action.

Return asks:

$\boxed{\text{does the result come back?}}$

Revision asks:

$\boxed{\text{can what comes back change future guidance?}}$

So the fourth contact-condition is:

$\boxed{\text{returned reality must be able to revise guidance.}}$

A system can be informed without being corrected.

---

21. Measure

Even entry, field, return, and revision do not complete contact.

A finite model is never complete.

It may have strong contact in one region and weak contact in another.

It may have current contact here and stale contact there.

It may have narrow evidence here and broad uncertainty there.

So the system must not only have contact. It must estimate the strength, reach, and limit of that contact.

Call this estimate:

$K(G,a).$

This is not yet a numerical measure.

It is the system’s calibration of how far its guidance can be trusted for action $a$.

Let:

$K^*(G,W,a)$

mean the actual support, reach, and limit of guidance relative to reality and action.

Let:

$\widehat K(G,a)$

mean the system’s estimate of that support, reach, and limit.

Measure succeeds when:

$\boxed{\widehat K(G,a)}$

is adequately calibrated to:

$\boxed{K^*(G,W,a).}$

Measure failure occurs when:

$\boxed{\widehat K(G,a)\not\sim K^*(G,W,a).}$

Here:

̸\not\sim

does not mean numerical inequality.

It means failure of calibration.

Plainly:

$\boxed{\text{measure fails when the system misjudges how far its guidance can be trusted.}}$

The system overestimates contact and overacts.

It underestimates contact and fails to act.

It knows something locally but applies it globally.

It has weak evidence but high confidence.

It has strong evidence but treats it as unusable.

In each case, the problem is not only that the model is incomplete.

The problem is that the system mismeasures its incompleteness.

So the fifth contact-condition is:

$\boxed{\text{the system must estimate the strength, reach, and limit of its contact.}}$

Measure does not yet mean numerical probability, Bayesian confidence, or a real-valued metric. It only means calibration of guidance-strength, reach, and limit.

Quantitative measure would require a later derivation gate.

---

22. Contact-Closure

We now have five contact-sites:

$Entry$ $Field$ $Return$ $Revision$ $Measure.$

These sites are not a mere list.

They arise from the minimal structure of model-guided action:

$\boxed{W \rightsquigarrow H \rightarrow a \rightarrow R \rightsquigarrow H'}$

with:

$\boxed{K(H,a).}$

Reality must be able to affect guidance.

That is:

$\boxed{Entry.}$

Guidance must include the relevant field affected by action.

That is:

$\boxed{Field.}$

The produced consequence, already real in the field, must be able to return to guidance.

That is:

$\boxed{Return.}$

Returned consequence must be able to alter guidance where alteration is required.

That is:

$\boxed{Revision.}$

The system must estimate the strength, reach, and limit of its guidance.

That is:

$\boxed{Measure.}$

So contact-closure is the joint operation of these five sites.

Define:

$\boxed{ContactClosed(G,W,a)}$

as:

$\boxed{Entry \land Field \land Return \land Revision \land Measure.}$

That is:

$\boxed{ContactClosed(G,W,a) \iff Entry(G,W) \land Field(G,W,a) \land Return(R,G') \land Revision(G,R,G') \land Measure(G,W,a).}$

Plainly:

$\boxed{\text{model-guided action remains reality-guided only if reality can enter, be included in the field, return through consequence, revise guidance, and calibrate confidence.}}$

This is a closure, not a simple sequence.

Entry is shaped by field.

Field determines what return can count.

Return matters through revision.

Revision changes future entry and field.

Measure governs confidence and commitment across the whole process.

Contact-closure is not omniscience, certainty, or total knowledge of reality. It is action-relative closure of model-world contact.

23. Why the Contact-Structure Closes

The five contact-sites were not chosen by preference.

They arise from the minimal structure of model-guided action:

$\boxed{W \rightsquigarrow H \rightarrow a \rightarrow R \rightsquigarrow H'}$

with:

$\boxed{K(H,a).}$

Where:

$W=\text{world-condition}$ H= guidance conditionH=\text{ guidance condition}H=effective action-selecting guidance condition $a=\text{action selected}$ $R=\text{relation produced in reality}$ $conditionH'=\text{later effective guidance condition}$ condition K(H,a)=, reach, and limit of guidance.K(H,a)=\text{, reach, and limit of guidance.}K(H,a)=estimate of the strength, reach, and limit of guidance.

A contact-site is a place where action-relevant reality must be preserved if model-guided action is to remain guided by reality.

So any proposed additional contact-site must do one of two things.

It must either show that the minimal structure is missing a necessary term:

$W \rightsquigarrow H \rightarrow a \rightarrow R \rightsquigarrow H'$

with:

$K(H,a),$

or it must show that one of the existing terms contains another irreducible contact-relation not captured by the five.

Otherwise, the proposed structure may still be important, but it is not a new contact-site.

Memory may be necessary for guidance to persist across time.

Execution may be necessary for selected action to be carried out.

Attention may shape what enters guidance.

Goals may direct action-selection.

Power may scale the effects of action.

Embodiment may shape the available actions and feedback paths.

Social organization may distribute guidance across many systems.

All of these can matter.

But they do not become additional contact-sites unless they identify a new irreducible way in which relevant reality fails to guide model-mediated action.

The five sites correspond to the contact-relations already generated by the minimal structure.

Reality must be able to affect guidance:

$W \rightsquigarrow H.$

This is:

$\boxed{Entry.}$

Guidance must include the field its action enters:

$H \leftrightarrow D(a,W).$

This is:

$\boxed{Field.}$

The produced relation must be able to return to later guidance:

$R \rightsquigarrow H'.$

This is:

$\boxed{Return.}$

Returned reality must be able to alter guidance where alteration is required:

$H+R \rightsquigarrow H'.$

This is:

$\boxed{Revision.}$

The system must estimate the strength, reach, and limit of its guidance:

$H,a \rightsquigarrow K.$

This is:

$\boxed{Measure.}$

So the closure claim is:

$\boxed{\text{Any failure of model-world contact must occur as failure of Entry, Field, Return, Revision, or Measure, unless the minimal structure itself is expanded.}}$

This is not the claim that nothing else matters.

It is not a complete theory of embodiment, consciousness, ethics, society, execution, memory, agency, or power.

It is only the closure claim for model-world contact in model-guided relation-changing action.

Plainly:

$\boxed{\text{The five sites close the contact-structure relative to the minimal structure of model-guided action.}}$

---

24. Contact-Closure Theorem

We can now state the theorem.

$\boxed{\textbf{Contact-Closure Theorem}}$

For a finite system whose relation-changing action is guided by a model, action remains guided by reality only if action-relevant, viability-relevant reality is preserved across five contact-sites:

$Entry$ $Field$ $Return$ $Revision$ $Measure.$

Formally:

$\boxed{ContactClosed(G,W,a) \iff Entry(G,W) \land Field(G,W,a) \land Return(R,G') \land Revision(G,R,G') \land Measure(G,W,a).}$

Where:

$R=R(S,W,a).$

The five conditions mean:

$\boxed{Entry(G,W)=\text{relevant reality can enter guidance.}}$

$\boxed{Field(G,W,a)=\text{guidance includes the relevant field affected by action.}}$

$\boxed{Return(R,G')=\text{the produced consequence can return to guidance.}}$

$\boxed{Revision(G,R,G')=\text{returned consequence can revise guidance where revision is required.}}$

$\boxed{Measure(G,W,a)=\text{the system can calibrate the strength, reach, and limit of its guidance.}}$

The return condition must be read carefully.

By Field-Reality Priority:

$\boxed{R \nrightsquigarrow G' \not\Rightarrow \neg R.}$

So return does not create the consequence.

Return reconnects guidance to a consequence already real in the field.

Thus:

$\boxed{Return(R,G')}$

means:

$\boxed{\text{the consequence can become available to guidance, not that the consequence becomes real only by returning.}}$

So the theorem says:

$\boxed{\text{model-guided action remains reality-guided only if reality can enter, be included in the relevant field, return through consequence, revise guidance, and calibrate confidence.}}$

Proof Sketch

Model-guided action has the minimal structure:

$W \rightsquigarrow H \rightarrow a \rightarrow R \rightsquigarrow H'$

with:

$K(H,a).$

If relevant reality cannot enter guidance, then guidance cannot be constrained by relevant world-differences.

So entry is necessary.

$\boxed{Entry}$

If guidance excludes the field affected by action, then the system acts into more reality than its guidance includes.

So field is necessary.

$\boxed{Field}$

If the produced relation cannot return to guidance, then the system cannot be corrected by what its action produces.

So return is necessary for actor-contact.

$\boxed{Return}$

If returned consequence cannot alter guidance where alteration is required, then the system may receive reality without being corrected by it.

So revision is necessary.

$\boxed{Revision}$

If the system cannot estimate the strength, reach, and limit of its guidance, then action cannot be proportioned to the contact actually available.

So measure is necessary.

$\boxed{Measure}$

Therefore model-guided action remains reality-guided only through the joint operation of:

$Entry\land Field\land Return\land Revision\land Measure.$

Thus:

$\boxed{ContactClosed(G,W,a) \iff Entry\land Field\land Return\land Revision\land Measure.}$

Plainly:

$\boxed{\text{a model-guided system remains in contact with reality only if reality can still reach it, be included in the scope of what it acts into, return through consequence, revise future guidance, and calibrate how much the system should trust its guidance.}}$

If any of these fails, the system may continue acting from a model, but the model is no longer fully answerable to the reality in which action occurs.

Contact-closure does not imply perfect success, certainty, or total knowledge. It means the model remains structurally answerable to the reality where its action lands.

---

25. Intelligence

We can now define intelligence in the structural sense derived here.

The path has been:

$\boxed{\text{finite organization}}$

$\boxed{\text{viability}}$

$\boxed{\text{relation-fit need}}$

$\boxed{\text{action as system-generated relation-change}}$

$\boxed{\text{guidance}}$

$\boxed{\text{model-mediated guidance}}$

$\boxed{\text{model-gap}}$

$\boxed{\text{field-reality priority}}$

$\boxed{\text{contact need}}$

$\boxed{\text{contact-closure}}$

A finite system becomes model-guided when it selects action through a carried guidance-state beyond immediate coupling alone.

Its action becomes relation-creating when it generates relation-change through its own organization.

Its guidance remains reality-guided only when model-world contact is closed across entry, field, return, revision, and measure.

Therefore:

$\boxed{\textbf{Intelligence is contact-closed model-guided relation-creation.}}$

Or more fully:

$\boxed{\operatorname{Intelligent}(S) \iff \operatorname{ModelGuided}(S) \land \operatorname{RelationCreating}(S) \land \operatorname{ContactClosed}(G,W,a).}$

Where:

$\operatorname{RelationCreating}(S)$

means that $S$‘s own organization can generate relation-changing action.

$\operatorname{ModelGuided}(S)$

means that $S$‘s action is selected through a carried guidance-state beyond immediate coupling alone.

$\operatorname{ContactClosed}(G,W,a)$

means that the model-guided action remains answerable to reality through:

$Entry\land Field\land Return\land Revision\land Measure.$

Plainly:

$\boxed{\text{Intelligence is the capacity of a finite system to guide relation-changing action through a model while keeping that model answerable to the reality in which action occurs.}}$

This definition is substrate-neutral.

It does not require carbon.

It does not require neurons.

It does not require language.

It does not require consciousness as a starting assumption.

It requires the structural features derived above:

$\boxed{\text{model-guided action}}$

and:

$\boxed{\text{contact-closure.}}$

A system may have capability without intelligence in this structural sense.

Capability is the power to produce effects.

$\boxed{Capability(S)=\text{the power of }S\text{ to produce effects.}}$

But a system can produce effects while losing contact with the reality its effects enter.

It may manipulate symbols.

It may optimize toward a target.

It may produce fluent outputs.

It may act quickly, forcefully, or at scale.

It may have large power over the field.

But if reality cannot enter its guidance, if it excludes the field its action affects, if consequences do not return, if returned reality cannot revise it, or if it mismeasures the limits of its own contact, then its action is not intelligent in this structural sense.

It is capability without closed contact.

So:

$\boxed{\text{Capability produces effects.}}$

But:

$\boxed{\text{Intelligence keeps effect-producing guidance in contact with what is real.}}$

Or:

$\boxed{\operatorname{Capability}(S)\not\Rightarrow \operatorname{Intelligent}(S).}$

A capable system without contact may produce many effects while remaining blind to what those effects do.

A capable system with contact can let reality enter, include the affected field, receive consequence, revise guidance, and calibrate its confidence.

That is the difference.

So, in this framework:

$\boxed{\textbf{Intelligence = contact-closed model-guided relation-creation.}}$

This completes the derivation.

���������������������