The parliament problem: why covenant alone is not enough
What fixes what fixes what governs?
Across domains as different as software, law, and personal behaviour, the same structural problem appears: systems fail not because they lack rules, but because they cannot sustain them under pressure. What presents as discipline, morality, or governance turns out to be an instance of a deeper pattern—one that defines the boundary between stability and collapse.
In the previous essays [one, two, three] I approached this through addiction—not as chemistry or desire, but as a failure of self-governance. The decisive moment is not the anticipated “high”, but the relief that arrives when the deeper dilemma of authority is resolved—prematurely. That insight led to a four-layer model of decision and authority, culminating in covenant: the structure that fixes what is allowed to decide.
Yet even covenant is not enough. Given sufficient pressure, what we treat as non-negotiable becomes negotiable again. The system of choice does not fail because it is weak, but because it can still revise what was meant to bind it.
So why does the parliament reconvene even after the king has spoken?
The parliament problem
We can now restate the structure that has been built across the previous essays.
At the surface is Layer 0 — behaviour: what happens. But behaviour does not explain itself.
Beneath it sits Layer 1 — authority: what decides what happens; every decision over behaviour already implies the selection of a deciding authority.
Under pressure, the system must determine which of these authorities is allowed to decide. This is Layer 2 — governance: what determines what counts as authority in the first place. Yet even governance alone does not hold. If what is allowed to decide remains open to revision, then when stressed the system will collapse into whatever resolves the tension fastest — another vodka.
And so we arrive at Layer 3 — covenant: the structure that fixes governance itself, removing the possibility of renegotiation at the point where it would otherwise fail. In theory this is where sovereignty enters; a “king” has been chosen to rule.
At this point, the system appears complete. Behaviour follows authority. Authority is stabilised by governance. Governance is fixed by covenant. The regress appears to have been closed.
And yet, this is not what happens.
Even where covenant is explicit—personal, moral, or institutional—collapse and relapse still occur. What was treated as non-negotiable becomes negotiable again. Not because of weakness, but because the system has shifted into a different operating mode: one in which it can no longer afford to sustain the full grounding that made the covenant binding — whether through baptism, membership of Alcoholics Anonymous, or legal commitments.
As load is increased, the cost of maintaining that covenantal grounding grows too high.
And when that happens, the system does not hold the line. It descends. What was fixed becomes open again. Governance returns to deliberation. Authority fragments into competing voices. The question of what should decide must be answered once more.
The parliament reconvenes — and with it returns the authority to question what was meant to be fixed.
Why self-authored covenant fails
We can now see why the problem persists:
If I set the rule, I retain the authority to unset it.
This is not a matter of intention, but of structure. Any covenant that originates from the self remains, in principle, subject to that same self. The act of binding does not remove authorship; it preserves it.
And so self-binding is, in fact, reversible binding.
Under ordinary conditions, this may not matter. The rule appears stable. It holds across time and guides behaviour. But applying pressure exposes the underlying structure. The same authority that established the rule reasserts itself — and with it the capacity to revise or suspend what was previously fixed.
At that point the parliament has already reconvened.
The rule no longer functions as covenant. It collapses into preference — one option among others, to be weighed against competing demands. This is not a moral failure. It is not a lack of discipline or sincerity. It is a consequence of recursion: the authority that sets the rule remains present at a higher level, able to reconsider what it has set.
And once that reconsideration is possible, it becomes inevitable under sufficient pressure.
The problem, then, is not that we fail to bind ourselves strongly enough. It is that self-authored binding cannot, by its nature, remove the authority that would need to be constrained.
The instability is structural.
What the “high” really is
We can now return to addiction, but with a clearer understanding of what is actually happening at the point of decision.
In the earlier essays I described the “high” as a synthetic governance object: a simplified construct that presents itself as sufficient to justify action, standing in for a much longer and more costly causal chain.
What matters is not the act itself, but the moment of commitment. The tension of whether to act collapses, and a sense of relief appears—before any substance is taken, before any reward has arrived.
This is often misinterpreted as anticipation of pleasure. But structurally, something else has occurred.
The system is no longer required to ask the deeper question of what should decide. The internal conflict is resolved, not by answering it at the appropriate level, but by terminating it. The need for further governance disappears.
At that moment the system experiences a form of closure—but at a level too shallow to sustain coherence over time.
This allows us to refine the model:
The “high” is not simply the anticipation of pleasure; it is the temporary removal of higher-order constraint—experienced as sovereignty.
Nothing stands above the decision. No further appeal is required. No competing authority needs to be reconciled. The system is, for that moment, self-authorising.
This is why the experience feels not only pleasurable, but right.
But this “rightness” does not come from alignment. It comes from the absence of contradiction. What has been removed is not the problem, but the requirement to answer it. The relief is real. The closure is genuine. But the structure that produced it cannot hold.
What appears as freedom is the suspension of governance.
The mechanism of pseudo-sovereignty
We can now describe the mechanism more precisely.
At the point of decision, multiple authorities are in play. Each presents a claim to decide — long-term wellbeing, immediate relief, prior commitments, external obligations. Under normal conditions these remain in tension.
In the moment of collapse, that tension is not resolved. It is eliminated. Competing authorities are dismissed, not defeated.
One construct is elevated and treated as sufficient to decide. The hierarchy collapses into a single line. No higher authority is consulted. No competing claim is allowed to stand. The question of what should decide has been closed by removing all but one candidate.
This produces a distinctive experience: calm, certainty, the sense that “this is right.”
But this is not the result of genuine alignment. It is the effect of removing opposition. What is felt as clarity is, in fact, the absence of contradiction. The system is no longer in conflict — not because the conflict has been resolved, but because it has been silenced.
This is pseudo-sovereignty.
The system experiences itself as unified and decisive, but only because the conditions for disagreement have been stripped away. What is taken as truth—sufficient to act—is simply what remains once nothing else is allowed to speak.
Why this is mistaken for truth
The confusion is structural.
The collapse of tension is taken as evidence of correctness. When the internal conflict disappears, the system interprets the resulting silence as proof that it has found the “right” answer. But these are not the same.
Tension collapse does not imply truth.
What the system actually registers is the absence of contradiction. The competing voices have been silenced. The noise has gone. The question appears settled. This is then misread as alignment.
“It feels right” is taken as “it is right.”
But what has occurred is not alignment; it is the suppression of dissent. The conditions required for disagreement have been removed, and the resulting silence is mistaken for coherence.
The distinction is critical:
Genuine alignment resolves competing authorities at the appropriate level.
Pseudo-sovereignty achieves the opposite: it eliminates competing authorities altogether, producing the appearance of unity without the underlying alignment.
Pseudo-covenant as authority substitution
We can now define what has been operating in place of covenant.
A pseudo-covenant is a standing resolution of authority via tension collapse. It does not fix governance in any durable sense, but it reliably produces the effect of closure: the parliamentary debate is ended, the question is settled, and action can proceed.
The function of a pseudo-covenant is not to establish truth, but to terminate deliberation:
Relief is substituted for truth.
Immediacy is substituted for legitimacy.
What matters is not whether the decision is grounded, but whether it resolves the tension quickly enough to allow the system to continue. This is why the mechanism is so compelling: it resolves tension immediately, forestalling the anticipated collapse that comes with perceived contradiction.
It requires no justification, because it bypasses the effort of self-governance entirely. And it is always available under load — the path of least resistance when higher-order processing can no longer be sustained.
Importantly, this does not mean that lower modes defeat higher ones. The system does not reject its own commitments or values in any principled sense. Rather, when those higher modes cannot be sustained, lower modes remain available as fallback. The system descends to what it can still execute when stressed.
Seen in this light, addiction is not a series of isolated failures. It is a loop of repeated authority substitution under load.
Each cycle reinstates the same pattern: tension emerges, governance cannot hold, pseudo-covenant resolves the question, and the system proceeds under a construct that cannot sustain it.
The loop persists not because it is chosen, but because it is available.
Why covenant still fails
Covenant is meant to do one thing above all: remove the possibility of renegotiation. It exists to fix governance — to close, once and for all, the question of what is allowed to decide.
But if the covenant is self-authored, it remains subject to revision.
The authority that establishes the rule is never truly removed by the act of binding. It remains present, and under pressure it reasserts itself. What was fixed can always be reopened, because the system retains the capacity to reconsider its own decisions.
This brings us back to the core problem.
The parliament remains sovereign over the covenant.
Not because it is explicitly granted that power, but because the system cannot sustain the conditions under which covenant can remain binding. When the cost of maintaining full grounding becomes too high, the system descends. Governance returns to deliberation.
The question that covenant was meant to terminate is thrown open once more.
At that point, covenant no longer functions as covenant. It becomes merely one position among others — subject to the very process it was designed to end. And once that process is active, the outcome is no longer in doubt.
The missing layer
We are now in a position to see what is missing — whatever leaves space for the inner parliament to sabotage our alignment to truth and covenant.
If covenant cannot hold when it remains subject to reinterpretation, internal override, or situational pressure, then something else is required — something not subject to these conditions at all.
Not a rule, which can be revised.
Not a decision, which can be reconsidered.
Not a preference, which can be overridden.
Its function is not to end the parliament by further decision, because assigning decision authority to ourselves is the originating problem.
Rather, it prevents the descent that would reopen its jurisdiction in the first place. The task is not to resolve the question of governance once it appears, but to maintain the conditions under which the question never needs to be reopened.
In other words, it keeps full grounding available even under pressure. That is the critical distinction.
Covenant attempts to fix what is allowed to decide.
This deeper layer determines whether that fixing can hold at all.
It is not something we decide.
It is what determines whether our decisions bind.
Reframing “worship”
We can now return to a term that is often misunderstood.
At Layer 3, covenant names the binding structure: what fixes governance and determines what is allowed to decide.
But this still leaves an open question: what fixes covenant?
Traditionally, this has been described as worship — not in the narrow sense of ritual or belief, but in a purely structural sense: what is ultimately served. This is not a theological claim. It is a structural one.
Seen in this light, worship is not what we say we value. It is what the system cannot subject to its own evaluation. It is what cannot be debated, cannot be overridden, and is not open to reconsideration when pressure is applied.
Worship is what gives covenant its binding force.
If what we serve remains open to revision, then covenant cannot hold. It will collapse back into governance, and governance back into deliberation and self-will.
But where something stands beyond that process — where it is not available to be questioned or displaced — the regress is halted.
In that sense, worship is not an optional layer added on top of the system.
Worship is what determines whether the system can stabilise at all.
True and pseudo covenant
We can now distinguish between what only appears to bind, and what actually does:
A pseudo-covenant is self-authored. It remains subject to revision, collapses under pressure, and requires ongoing enforcement to sustain it. It must be continually reasserted because the authority that established it remains able to reconsider it.
A true covenant operates differently. It is anchored beyond the system that would otherwise revise it. It does not depend on continuous enforcement, because it does not remain open to reconsideration. Even under extreme cost, it keeps full grounding available — and in doing so stabilises governance without ongoing effort.
The contrast is not moral, but structural:
A pseudo-covenant must be maintained.
A true covenant holds.
This explains why pseudo-covenants persist. They resolve tension faster than truth can be established. They require less capacity. They are always available under load. When the system cannot sustain full grounding, they provide an immediate substitute.
Under these conditions, the outcome is predictable:
The system defaults to pseudo-covenant under load.
This is the mechanism of the addictive loop.
Resolution of the parliament problem
The parliament dissolves when it no longer has authority to legislate.
This does not occur through better arguments, stronger reasoning, or greater willpower. The problem is not solved at the level of deliberation.
It is solved by the loss of jurisdiction.
When the parliament is no longer sovereign — when it no longer has authority over what may be questioned — it cannot reopen what has been fixed. The process of debate does not conclude. It becomes irrelevant. This is the critical shift.
The question is not settled by winning the argument, but by removing the authority of the forum in which it takes place.
In structural terms, the parliament dissolves when descent is no longer possible. When the system can no longer fall back into modes where governance must be renegotiated, the conditions for deliberation do not arise.
Without descent, there is no reopening.
Without reopening, there is no parliament.
And without parliament, covenant holds.
The institutional mirror
The same structure does not remain confined to the individual. It recurs in law.
A system that cannot stabilise what counts as authority does not stop functioning. It continues — but on a different basis. Instead of resolving the question of authority, it treats the process itself as sufficient to proceed.
In effect, the procedural parliament becomes sovereign.
What should be subordinate to statutory or constitutional covenant — what fixes what is allowed to decide — is instead reopened and managed through ongoing process. The question of authority is not settled. It is displaced.
The result is a familiar pattern:
Authority is asserted without formal grounding or evidential support.
Contradiction is not resolved, but managed through continuation.
Proceedings advance, decisions are issued, and enforcement follows.
The appearance of legitimacy is preserved, even as its basis remains unsettled. This is not a failure of individuals. It is a structural response to a system that cannot hold its own conditions of authority.
The parallel with addiction is exact.
The addict resolves tension through personal behaviour.
The institution resolves tension through process—its collective behaviour.
In both cases, the deeper question is not answered. It is bypassed.
And in both cases, what sustains the system in the moment undermines it over time.
Return to addiction
We can now return to addiction with a clearer understanding of what is at stake.
The problem is not a lack of knowledge. Those caught in these loops are typically well aware of the consequences.
Nor is it a lack of rules. Commitments are made, lines are drawn, and intentions are often sincere.
The problem is that those rules remain debatable.
Under pressure, the system descends to a level at which the parliament must exist. The question of what should decide reopens, and what was meant to bind becomes one option among others.
At that point, the outcome is no longer determined by what is known or intended, but by what can resolve the tension fastest.
This is why addiction persists.
It is not an attempt to end the self. It is an attempt to continue without resolving contradiction.
The system does not seek destruction. It seeks relief — any resolution that allows it to proceed.
And this is why the “high” wins.
Not because it is more pleasurable, nor because it is believed to be good, but because it ends the question — temporarily — while preserving the system.
It allows the process to continue without confronting what cannot be held.
A practical bridge
This model suggests a simple extension to the earlier intervention.
Before you decide, ask the familiar question:
“If this fully plays out, where does it leave me?”
This shifts attention from immediate reward to downstream state — from income to capital.
But this alone is not sufficient.
A second question is required:
“What ultimately anchors this decision so the parliament cannot reopen it tomorrow?”
The first question restores attribution.
The second tests whether that attribution can hold.
Together, they do not remove the decision. They change the conditions under which the decision even arises.
Final synthesis
We can now restate the structure in full.
Layer 0 — behaviour: what happens.
Layer 1 — authority: what decides what happens.
Layer 2 — governance: what determines what is allowed to decide.
Layer 3 — covenant: what fixes governance.
Layer 4 — worship: what gives covenant authority at all.
Each layer exists because the one above it cannot stabilise on its own. Each attempts to resolve a deeper question that the previous layer cannot answer.
The pathology arises when this structure is short-circuited. Higher-order contradictions are resolved at lower layers. Questions that belong at the level of governance or covenant are terminated at the level of authority or behaviour. The system does not answer them. It closes them prematurely.
Or more simply: The question is ended at a level too low for the decision being made. This is what we have seen in addiction, in institutions, and in ourselves.
And it leads to a final conclusion:
As long as we remain the highest authority, nothing can bind us.
In closing
Covenant alone is not enough.
The problem is not drink, drugs, or debauchery.
In institutions, it is not delay, dysfunction, or disorder.
The problem is that everything remains open to debate.
As long as the question remains open, it reopens.
As long as it can be debated, it can be overridden.
The real question is:
What fixes what fixes?
Answer that, and there is nothing left to debate.