Saturday, 17 January 2026

Misalignment: 3 Why Correction Fails

When misalignment becomes visible, the response is almost always corrective. Something must be added, removed, clarified, or improved. More information will close the gap. Better models will resolve the tension. Clearer explanations will bring meanings back into alignment. Where persuasion fails, optimisation is invoked: incentives adjusted, behaviours nudged, systems tuned.

These responses are not irrational. They are locally effective, often indispensable, and deeply sedimented in our practices. But they share a common assumption: that misalignment is a problem of insufficiency. Something is missing that, if supplied, would restore coherence.

This post argues that this assumption is false. Correction fails not because it is poorly executed, but because it targets the wrong level. Misalignment is not a deficit within a construal, but a relation between construals that do not compose.


The Additive Fantasy

Correction typically takes an additive form. We assume that if one construal fails to align with another, it is because it lacks some relevant element: a fact, a variable, a perspective, a piece of context. The remedy is therefore accumulation. Add what is missing, and alignment will follow.

This fantasy is sustained by the success of additive strategies in bounded domains. Within a stable construal regime, additional information can indeed refine judgements and improve coordination. But this success does not generalise. When construals fail to compose, adding material to one does not repair the relation between them. It often exacerbates it.

More information does not guarantee compatibility. It can sharpen distinctions that make composition even less admissible.


Explanation as Displacement

Explanation is often treated as a neutral good. To explain is to illuminate, to render intelligible, to bridge gaps in understanding. In situations of misalignment, explanation is deployed as a primary tool of repair.

But explanation always operates within a construal. To explain something is to redescribe it in terms that are already meaningful within a given regime. When misalignment is present, explanation does not travel freely across the gap; it displaces the problem into one construal’s terms.

This displacement is easily mistaken for resolution. The explainer experiences increased coherence; the explained-to may experience pressure to translate, accommodate, or defer. The underlying failure of composition remains untouched.


Optimisation Without Meaning

Where explanation falters, optimisation often takes its place. If meanings cannot be aligned, perhaps behaviours can be. Incentive structures, feedback loops, and performance metrics are adjusted to produce coordination without requiring shared construal.

Optimisation can be effective at the level of outcomes. It can stabilise interaction and suppress overt conflict. But it does so by bypassing meaning rather than repairing it. Coordination achieved in this way is brittle. It depends on continued enforcement and breaks down when conditions shift.

Optimisation treats misalignment as noise to be managed, not as a structural condition to be understood. It trades intelligibility for control.


The Appeal to Reality

Another corrective reflex is the appeal to reality itself. Faced with incompatible construals, one may insist that “the facts decide,” that “reality will correct us,” or that “the world does not care about our interpretations.”

Such appeals presume that reality presents itself independently of construal, ready to arbitrate disputes. But there is no access to reality that is not already mediated by distinctions that make it intelligible. Invoking reality introduces yet another construal, whose compatibility with others is not guaranteed.

Appeals to reality do not resolve misalignment; they merely assert the authority of one construal over others.


Why Correction Persists

If correction fails so systematically, why does it persist as the default response? Part of the answer lies in habit and institutional inertia. Corrective practices are built into education, governance, and expertise. They work well enough often enough to justify their extension.

More deeply, correction persists because it preserves the fantasy of global coherence. It allows us to believe that misalignment is temporary, local, and ultimately eliminable. To abandon correction as a general solution is to accept that alignment is costly, partial, and contingent.


What Comes After Correction

Rejecting correction does not entail passivity or relativism. It entails a shift in what we take the problem to be. If misalignment is not a deficit to be repaired, then the task is not to fix meanings, but to examine how construals interact, interfere, or fail to relate.

In the next post, we will turn to a consequence of this shift that is often resisted: the possibility that misalignment can persist without breakdown, and that stability itself may depend on non-alignment.

No comments:

Post a Comment