Aliases: Untools; abstraction laddering; concept map; issue tree; ladder of inference; second-order thinking; inversion; reasoning.
| Uncertainty | Lens | Useful Output | Limit |
|---|---|---|---|
| Request assumes a solution | Abstraction laddering | Purpose and bounded problem before mechanism | Moving upward does not authorize scope expansion. |
| Terms/relationships are confused | Concept map | Explicit subject–relationship–object statements | A drawn edge is not an implemented contract. |
| Several causes are plausible | Issue tree | Nonoverlapping investigation branches within scope | Do not call an unexamined checklist exhaustive. |
| Observation becomes a confident claim | Ladder of inference | Separate observation, interpretation and assumption | Timeout alone cannot establish non-effect. |
| Local change affects other work | Second-order thinking | Consumer and outstanding-work consequences | Stop speculative expansion beyond the relevant change. |
| A guarantee might be bypassed | Inversion | Concrete invalid-use/counterexample candidates | The project decides which evidence to maintain. |
Example: “add an approved flag” is a representation request. First recover the commitment: publish the exact reviewed version. Then identify ownership, required current facts and a mechanism that retains binding. Inversion tests whether a caller can forge true or substitute a version. None of these tools grants authority to change who approves.
These tools are selected aids, not a universal six-step process. Keep original requirements and consequence reasoning visible when comparing designs. Source explanations and original worked applications are in the doctrine's thinking-tools reference, linked to Untools.
Related: CS-02, CS-03, CS-05, CS-06, CS-09, CS-12, CS-15, CS-17.