Bow Tie Kreative Writing System
Run the engine

Rules / clarity

CLR-020 Hard constraint layer: paragraph

Separate fact, inference, and opinion

IF a paragraph contains facts, inferences, or opinions, THEN mark each proposition’s epistemic status when confusion would affect trust or decisions ELSE retain.

Why the rule exists

Readers should know what is observed, concluded, believed, or forecast.

Condition

The rule fires when this evaluates true. Paths marked below must be present in the fact base; the engine reports the rest rather than guessing them.

  • ALL
    • paragraph.epistemic_type_count gt 1
    • paragraph.status_confusion_consequential eq true

Facts this rule reads

PathSource
paragraph.epistemic_type_count You supply it
paragraph.status_confusion_consequential You supply it

Then

  1. label — target paragraph.propositions
  2. cite — target documented_facts

Else

  1. retain — target current_text, No triggering condition is present.

LAKA

Smallest sufficient intervention
Yes
Primary axes

Sources

  • Technical Communication — Mike Markel; Stuart A. Selber
    Audience, design, ethics, research, and technical genres.
  • Advertising and Marketing Basics — U.S. Federal Trade Commission link
    Advertising claims must be truthful, non-deceptive, fair, and appropriately substantiated.
  • Advertising Substantiation Policy Statement — U.S. Federal Trade Commission link
    Objective claims require a reasonable evidentiary basis before dissemination.
The raw rule record

Verbatim from 11-clarity-style-rules.json. Also available at /v1/rules/CLR-020.

{
  "id": "CLR-020",
  "name": "Separate fact, inference, and opinion",
  "domain": "clarity",
  "layer": "paragraph",
  "strength": "hard_constraint",
  "human_logic": "IF a paragraph contains facts, inferences, or opinions, THEN mark each proposition’s epistemic status when confusion would affect trust or decisions ELSE retain.",
  "when": {
    "all": [
      {
        "path": "paragraph.epistemic_type_count",
        "operator": "gt",
        "value": 1
      },
      {
        "path": "paragraph.status_confusion_consequential",
        "operator": "eq",
        "value": true
      }
    ]
  },
  "then": [
    {
      "action": "label",
      "target": "paragraph.propositions",
      "labels": [
        "documented_fact",
        "inference",
        "opinion",
        "prediction"
      ]
    },
    {
      "action": "cite",
      "target": "documented_facts"
    }
  ],
  "else": [
    {
      "action": "retain",
      "target": "current_text",
      "reason": "No triggering condition is present."
    }
  ],
  "unless": [],
  "because": "Readers should know what is observed, concluded, believed, or forecast.",
  "diagnostics": [],
  "examples": {},
  "source_ids": [
    "BK-026",
    "WEB-012",
    "WEB-013"
  ],
  "laka": {
    "smallest_sufficient_intervention": true,
    "primary_axes": [
      "scope",
      "depth",
      "magnitude",
      "reversibility"
    ]
  }
}

Other clarity rules

  1. CLR-001 Prefer the precise familiar word IF two words are equally accurate AND one is more familiar to the audience, THEN use the familiar word ELSE keep the precision-bearing term. Context default
  2. CLR-002 Define necessary jargon IF a domain term is necessary AND the intended reader may not know it, THEN define it at first use and use it consistently ELSE replace it with plain language. Context default
  3. CLR-003 Expand unfamiliar abbreviations IF an abbreviation is not universally familiar to the audience, THEN spell it out at first use and pair it with the abbreviation ELSE retain. Context default
  4. CLR-004 One term per concept IF multiple terms refer to the same technical concept without rhetorical need, THEN select one preferred term and replace the variants ELSE retain. Context default
  5. CLR-005 One concept per term IF the same term denotes multiple concepts in one context, THEN rename or qualify the concepts ELSE retain. Hard constraint
  6. CLR-006 Replace vague reference words IF THIS, THAT, IT, THEY, FORMER, or LATTER has an uncertain referent, THEN repeat or summarize the referent ELSE retain. Hard constraint
  7. CLR-007 Quantify actionable vagueness IF vague quantity, frequency, time, or probability affects a decision, THEN provide a range, threshold, date, or defined category ELSE retain approximate language. Context default
  8. CLR-008 Use concrete examples after abstractions IF a central abstraction lacks a concrete instance AND the audience knowledge gap is moderate or large, THEN add one representative example ELSE retain. Heuristic