{"path":"research/grounding/scope.read.md","content":"What this project has ruled that bears on whether one claim speaks of one case or of many, and if many, whether it is a strict, normic or statistical generalization, in plain words, so you know how careful the question is before you read the item:\n- A generalization carries one number, how sure we are that it holds as written; its usually is a scope word, made exact in its wording: a normic generalization, whose exception calls for an explanation, becomes the universal plus named exceptions (the gate is open every Tuesday plus it is closed on public holidays).\n- Every claim's scope is read: every claim, wherever it came from, is read once by a standing pass: does it speak of one case or of many, and if many, is it a strict, normic or statistical generalization, or a generic whose words do not settle its kind? Code reads what the words state (every, always, a stated proportion, a named case at a time) and a model reads the rest, with one case or many? asked first as a gate whose no-answer is cannot tell. The reading is dated and versioned, read and never set, and a person disputes it by contributing a rival wording. The founder's words: \"I think all of these sound exciting, and I want to see where we can go with this. I think we're onto something huge. I have some reflections on it, but bless all these now, and then we can take it from there.\".\n- A strict generalization has no exceptions: a rule is strict when a failure cannot be explained without the rule being wrong (a definition, a theorem, every bachelor is unmarried): a counterexample refutes it, and the repair is a narrower rival wording. Strict, normic and statistical are the three kinds of rule, told apart by two questions about a failed case: does it call for an explanation (no: statistical), and could an explanation leave the rule standing (no: strict; yes: normic). A claim about one case is no rule. The founder's words: \"I think all of these sound exciting, and I want to see where we can go with this. I think we're onto something huge. I have some reflections on it, but bless all these now, and then we can take it from there.\".\n- An exception is a claim of its own: no claim carries an ability to have exceptions. Any claim can receive a counterexample as an objection, and any objection can have its relevance attacked by an explanation; both are ordinary moves available everywhere, and nothing about them is stored on the claim. The claim's scope reading decides what the strength arithmetic does with them: for one case the counterexample is the claim's denial, for a strict generalization it refutes, for a normic generalization it counts until explained, for a statistical generalization it moves the rule a little, either way, as every case does (revised 2026-10-23 from counts only through frequency; what-a-case-counts-for.md § 5). Exception names one outcome: an explained counterexample to a normic generalization. The founder's words: \"I think all of these sound exciting, and I want to see where we can go with this. I think we're onto something huge. I have some reflections on it, but bless all these now, and then we can take it from there.\".\n- A generalization, and its role as a rule: a claim that speaks of many cases is a generalization (strict, normic or statistical); rule names what a generalization does under a link, the general sentence an inference applies. The founder's words: \"Ruled, make it so.\".\n"}