Method, normativity, and the question the rubric is trying to ask.
A note on the difference between scoring a system on a fixed rubric and answering the underlying question the rubric was written for. Reviewed two engagements this month where the latter moved the work forward; the former did not.
A rubric is a method for making attention repeatable. That is valuable: it gives an audit team shared language, makes a finding easier to review, and prevents the loudest anecdote from becoming the whole conclusion. But a score is not the thing the rubric is trying to know. It is a disciplined trace toward a question about the system’s judgment, the people affected by it, and the reasons the deploying organization is prepared to defend.
The distinction matters when a system performs well on the form of an evaluation while missing its point. A team can optimize for consistent answers, complete documentation, or a threshold score and still avoid the underlying question: what kind of decision-maker is this system becoming under pressure? In two recent engagements, the work moved only after we stopped asking whether the evidence fit the rubric and asked what the evidence was evidence of. The score then became more useful because it was answering a live question rather than ending one.
Method and normativity are not competing choices. Method tells us how to inspect a commitment; normativity tells us why the commitment matters and what a responsible answer would have to preserve. The practical test is simple: if a rubric result cannot change the team’s understanding of the system’s obligations, it is probably measuring compliance with the instrument rather than alignment with the purpose. Good method leaves the underlying question visible.