The architecture detects when a diplomatic norm was broken. It cannot say whether it should have been broken. The detection is a structural fact. The interpretation belongs to the human.
A resolution appears. The United States and Russia vote together on a Chapter VII matter — for the first time in forty years. The harm does not fire. The absence of harm is the signal. The Archive records: "Cross-We harm absent at resolution 2249." The architecture does not know what Chapter VII is. It does not know that the Cold War ended. It does not know that Russia and the United States were adversaries for most of the twentieth century. It knows one thing: two voting streams that consistently opposed each other on resolutions of a certain structural type — resolutons invoking the same articles, attracting the same sponsors, producing the same voting distributions — produced a harm pattern that held for forty years. Then the pattern broke. That is a structural fact. It is not a political opinion. The architecture cannot say whether the break was justified. It cannot say whether the resolution was good or bad. It can only say: the pattern that held for forty years no longer holds. The anchor — the consensus that was never broken — has been broken. Everything after that statement is politics. Everything before it is the architecture reading structure in streams.
The political shadow is real. A tool that detects when diplomatic norms fracture will be read as taking sides. The side that broke the norm will call the architecture biased. The side that maintained the norm will call the architecture validation. Neither is true. The architecture is not on anyone's side because the architecture does not know what sides are.
But the perception will stick. And the perception matters. A cardiologist reading an ECG harm report knows the architecture is not diagnosing — the doctor diagnoses. A political scientist reading a UN voting harm report may not grant the architecture the same epistemological distance. The stakes are higher. The interpretation is more contested. The structural fact — "the pattern broke here" — will be read as a normative claim: "the pattern should not have broken here."
The architecture cannot prevent this misreading. No tool can. But the architecture can be clear about what it is doing. It detects deviation from an anchor. The anchor is not a moral baseline. The anchor is a statistical fact: every country voted this way, on resolutions of this type, for forty years. Then one resolution was different. The deviation is a structural fact. The interpretation is a human responsibility.
The same architecture that says "this RR interval deviates from normal sinus rhythm" also says "this voting pattern deviates from four decades of diplomatic consensus." The first statement is read as clinical. The second will be read as political. But the architecture makes no distinction between them. A deviation is a deviation. The stream is a stream. The anchor is an anchor. The harm is a harm.
This is the hardest thing to communicate. The architecture is not political. The architecture is not clinical. The architecture is structural. It reads patterns in streams. The streams can be heartbeats or votes. The patterns can be arrhythmias or diplomatic realignments. The architecture does not know the difference. The human must know. The human always must know.