The Telos Gap

I. The Visible Failure Nobody Owns

Some failures are hidden. The dangerous class is visible but ownerless.

A telos gap occurs when a real, mechanism-produced harm does not intersect with any actor’s effective purpose-structure. No institution is mandated, rewarded, authorized, or narratively positioned to repair it. The failure exists in reality, but it never becomes a public object of governance. Reports describe it. Surveys measure it. Experts warn about it. The political system processes around it because no one is forced to process it.

Here telos names the observable purpose-structure of an institution. An institution’s telos is the intersection of seven elements: (1) statutory mandate, (2) budget line, (3) measurable performance indicator, (4) legal powers, (5) internal incentive (career, reputation, electoral), (6) constituency that experiences gain or loss, (7) procedural trigger that puts a question on the agenda. A telos gap is what falls outside that intersection across all relevant institutions.

The remedy is not awareness. The remedy is objecthood under procedural constraint: the failure must be turned into something the political system is forced to accept, dispute, assign, or consciously tolerate.

The concept is falsifiable. If a harm has a mandated owner, a budget line, a KPI, a procedural trigger, and that owner is required to act or report within a defined review cycle, there is no telos gap. There may still be policy disagreement, fiscal failure, or capture — but those are different objects, owned by named actors. A telos gap is the case where disagreement can exist but repair ownership does not.

II. Purpose-Structures: How Institutions Decide What Counts

Institutions do not respond to reality directly. They respond to reality after it has been translated into their admissible objects.

A ministry sees a bill, a programme item, a budget frame, a statutory duty. A municipality sees local services, local voters, a local tax base. A firm sees revenue, liability, margin, regulation. A journalist sees conflict, novelty, scandal, personified agency. An academic sees publishable units. A party sees electoral advantage and coalition constraints. Each set of admissible objects is the product of long institutional adaptation; each is locally rational; each excludes most of reality from action.

A system-level harm can be known and still fail to enter any of these object-languages. Every actor is doing its job. The system is still losing what no actor is positioned to defend. This is a structural claim about the intersection of their object-spaces, not a moral claim about the actors.

The default explanations — incompetence, corruption, complexity — often mistake the absence of an owner for a property of the problem. Incompetence locates the failure inside an existing actor’s competence. Corruption locates it inside an existing actor’s incentives. Complexity locates it inside the harm itself. None of the three notices the prior step: the actor whose competence, incentives, or grasp of complexity would matter does not exist as an institutional object at all. The harm is not necessarily unsolvable; it is unowned.

The implication runs the other way too. An owner can be absent without anyone being incompetent or corrupt. Well-functioning bureaucracies produce telos gaps as a normal product of their division of labor. The gap is what falls between them, not what falls inside any of them.

III. Where This Is Not Just Relabeling

The hostile reader’s first move is to point at a neighboring concept and ask whether the new term adds anything. The honest answer is to spell out the residual.

The contribution is the slicing of one specific failure mode that the neighbors describe parts of without naming directly: the intersection failure across an institution’s admissible-object dimensions. The shorter operational name is repair-side ownership failure. A harm can be perceived, understandable, and politically uncontested, and still fall through every owner because no actor holds the combination of mandate, budget, metric, power, constituency, and trigger needed to make repair an ordinary duty. Luhmann describes how subsystems close. Scott describes how categories see. Telos gap describes what happens after both subsystems are open and categories are clear: the harm has a perceiver and no claimant.

The closest framings outside the academic table are Yudkowsky’s inadequate equilibria and Alexander’s Moloch: synthetic vocabularies for civilizational coordination failure. The telos gap is one named subtype inside that family — the case where the failure does not require active competitive pressure or strong-actor suppression but only the absence of repair-side ownership across the existing intersection.

One caveat. Some apparent telos gaps are manufactured by power. When an actor benefits from keeping a harm formally unowned, the operative mechanism is capture or agenda denial, not passive ownerlessness. The telos-gap residual is the narrower case: the harm fails to become an object even when no one is suppressing it. Capture is a different problem, with a different repair, owned by a different vocabulary.

IV. Qualification Test and Types

The concept needs a gate. Without one it absorbs every public complaint and becomes infinite.

A failure qualifies as a telos gap if all seven hold:

1. The harm is real or credibly forecast.
2. The harm is mechanism-produced.
3. The mechanism is describable through actors and incentives.
4. Repair is not effectively owned by any existing institution.
5. The ownership failure is topological: the relevant mandate, budget, metric, power, constituency, and trigger do not converge in one actor.
6. Evidence and falsification conditions exist.
7. Plausible mechanism paths can be examined.

The fifth criterion measures structure, not head count. The signature of a telos gap is non-convergence: the mandate sits in one place, the budget in another, the KPI in a third, the constituency in a fourth, and no actor holds enough of them to make the repair an ordinary duty.

Six recurring types:

The cleanest test case of the six is the hidden-adjustment gap. The official target is met because the system absorbs contradiction through variables no actor’s telos contains. A dashboard can say “on track” while queues, debt, staff turnover, quality decline, or informal rationing say otherwise. The dashboards are honest; they are scoped to the wrong object.

V. Two Specimens

A telos gap should be visible inside a specific case — visible enough that the orthodox alternative explanation can be tested against it and fall short.

Specimen 1: Infrastructure maintenance debt

The visible surface is decaying transport infrastructure — roads, railways, bridges, waterways — with deferred maintenance lines in budgets every year and visible failures over time. The hidden mechanism is that annual budgets prefer projects over trajectories, and no actor’s KPI or constituency includes the asset’s decay curve as a primary object. The failure mode is accumulated fragility until a visible event forces attention: a water main bursts, a bridge closes, a school is condemned.

The standard explanation is public choice: politicians prefer ribbon-cuttings to invisible maintenance because elections reward visibility. That is part of the picture. The other part shows up even where the technical object exists. Finland’s Transport Infrastructure Agency calculates and publishes the maintenance backlog annually (about €4.18 billion in 2025), and the state budget has periodically included dedicated allocations for backlog reduction. The decay trajectory has a number. What it does not have is the rest of the institutional intersection: a protected budget line that survives across electoral cycles, a binding trajectory target with consequences for missing it, a constituency that experiences the curve as present loss, and a procedural trigger that forces action when the curve crosses a threshold. Beyond underfunding, the result is partial objecthood — engineering knowledge, a published number, advisory standing, and no binding ownership of the repair trajectory.

The telos-gap reading points at what would have to be added to repair the gap. Not “more money for infrastructure” — that gets absorbed by the same project bias, sometimes by the same one-off allocations. A trajectory custodian: an actor whose statutory duty is the curve itself, with a binding trajectory target and a procedural trigger that fires when condition crosses a threshold rather than after a visible event. The gap survives political alignment; it is structural, not motivational.

Specimen 2: Care-system hidden adjustment

The visible surface is a public care system — Finland’s post-2023 wellbeing-services system is the local case — where individual indicators (wait-time breaches against the care guarantee, staffing shortages, county deficits, regional access disparities, unmet medical need) are documented across separate institutional channels: THL national assessments, county-level audits, supervisory findings, EU comparative reports. Each piece is visible. The telos-gap question is whether any actor owns the joint vector (queue length, staff stock, deferred care, service quality, regional access, financial pressure) as one equilibrium failure rather than as separate items in separate channels.

The standard explanation is Goodhart’s Law: when a metric becomes a target, it stops being a good measure, and unmeasured variables degrade. That explains the distortion. The telos-gap layer appears next: even after the joint variables are partially measured and reported, no actor’s admissible objects include the vector as one equilibrium failure. Each institution that touches a piece processes its piece correctly; the equilibrium falls between them.

The hidden adjustment is not exhausted by misaligned agents responding to wrong incentives. The telos-gap layer appears when the degraded joint vector remains unowned even after the pieces are partially measured. The repair would not be better metrics — those still get parsed by the existing institutions — but a new owner whose KPI is the joint vector itself, with a procedural trigger to escalate when the vector moves. Without that, the system continues to report compliance while absorbing the failure through variables that belong to no one.

VI. Knowledge Does Not Fix Non-Objecthood

A warning can be true and still become residue.

Optionality Has No Router describes the case where a corrective signal exists but cannot reach the actor with authority to use it. The architecture has the right roles; the wiring is broken. The telos gap is the prior failure mode: no actor exists in the architecture with authority to receive the signal, so better routing does not help.

The mechanics inside an unrouted information environment are familiar. An expert report enters the public sphere. A ministry treats it as one view among many. A journalist lacks a conflict frame. A party sees no electoral upside in adoption. A committee has no procedural trigger to consider it. A budget process has no line item to absorb it. The finding remains true and inert.

This describes what happens to information when the object it describes does not exist as a public object of repair, not a critique of expertise or the press. The information enters a system that has been optimized to ingest other kinds of information. The press optimizes for conflict and narrative. Politics optimizes for electoral salience. Bureaucracy optimizes for procedural fit. None of them is optimized for “true claim about a mechanism nobody owns.” The information does not get rejected. It gets metabolized as background.

Routing failure assumes a destination. Telos gap is the absence of a destination. Knowledge without objecthood becomes residue.

VII. Restoring Objecthood

The response to a telos gap is an institutional procedure that produces objects with properties knowledge alone cannot have. A public report by itself is not enough; reports get metabolized as background.

A public object of repair specifies, at minimum, seven things:

1. visible phenomenon
2. mechanism producing it
3. actors and incentives
4. equilibrium
5. evidence and uncertainty
6. ownership status — which institution owns the repair; if none, the formal ownerlessness is part of the object
7. mechanism paths

Specifying the object is necessary and not sufficient. A document with all seven properties can still get metabolized as background if no procedural pressure ingests it. The transmission mechanism is the second half of the repair.

The institution producing such objects must hold three procedural properties:

1. Versioned publication. The object exists with a date, a version history, an owner status, and a review deadline. It is not a one-time report.
2. Right of reply with deadline. Named institutions are formally notified and have a bounded period to accept ownership, dispute the mechanism claim, identify a different owner, or record conscious non-action.
3. Visible audit trail. The political system can no longer pretend the question never landed. The record of who responded and how is part of the object.

These three properties together force the system to do something with the object even if that something is conscious refusal. The default option — silence — is taken off the table. The institutional ground rule is no longer “knowledge enters the public sphere and dissolves” but “knowledge enters the procedure and the procedure produces a record of where it landed.”

This is not a transfer of authority over values. The procedure does not give experts the power to choose. It compresses the time between mechanism-claim and political response. The trade-offs that the procedure forces into view remain for elected actors to resolve, dispute, or consciously ignore. The power exercised is procedural, not substantive: it is the power to require that the question be answered, not to determine the answer.

The democratic constraint follows from the same observation. The procedure targets one narrow class — system-level harms that current purpose-structures systematically fail to register — and raises the procedural cost of pretending they do not exist. The ambition stops there. Beyond that boundary, the procedure has nothing to say about substantive aims.

One predictable objection: any new institution can develop its own telos gaps. The same intersection failure that ownerless harms exhibit could afflict an authority that publishes mechanism objects. The procedural design reduces that risk rather than eliminating it. Such an authority can be evaluated first by procedural compliance — did it publish the objects, did it route the responses, did it keep the audit trail — and second by external review of whether its objects later proved useful, misleading, or inert. Its outputs sit inside the same audit trail it produces. The displacement is not perfect. The class of failures becomes narrower and more tractable.

VIII. The Failures That Belong to Nobody

Modern governance produces failures that belong to nobody because it divides reality into mandates, budgets, metrics, jurisdictions, and narratives. What falls between them remains real and becomes ownerless. Knowledge alone leaves it inert. Procedure can move it. A civilization needs more than better answers. It needs institutions that make the right questions unable to disappear.

See also: Optionality Has No Router (when the destination exists but the wiring is broken) · The Unpopulated Meta (the missing-architecture case in detail) · Full-Stack Survival (the unmonitored capital layers as a special case of capacity-stock gap) · Telocracy · Mechanism Space