Attention & signal

Noise

Šum
Canonical MonkOS term

An internal or external impulse that disperses attention, breaks continuity, or imitates signal without sufficient coherence.

Noise is not merely complexity; it is dispersion that prevents a useful relation from carrying through.

Why this term exists

Complex systems can be rich without being noisy. The distinction matters because aggressive simplification can destroy useful structure while leaving the real source of dispersion untouched.

Definition in the system

Noise can be random, repetitive, manipulative or contextually irrelevant. It can also arise from excessive mediation: too many layers between a difference and the response it should inform.

Whether something is noise depends partly on context. A detail irrelevant to one task may be critical to another. MonkOS therefore treats noise reduction as selective rather than universal deletion.

The goal is a better signal-to-noise relation, not a sterile system. Tension, difference and Detraction can remain productive while unnecessary dispersion is reduced.

Example

A dashboard with twenty live metrics may contain more data than a dashboard with five, yet provide less signal if the operator cannot see which change affects the current decision.

What it is not

Noise is not synonymous with disagreement, complexity, novelty or Detraction.

Formal / operational note

Working normalized noise quantities are used in some MonkOS models, including a dimensionless dis-coherence index D and noise-temperature index θ.