A stated position is evidence. An inferred interest is a hypothesis.

The distinction is useful because it opens a question

In a fictional port dispute, an operator says, “The inspection window must remain at night.” It is tempting to label the interest as uninterrupted trade. That may be part of the explanation. The operator may also be concerned about staffing, a contractual delivery window, or the cost of changing a routine. The demand alone does not tell us which explanation matters, whether several matter, or whether the speaker is authorized to describe the organization’s priorities.

Harvard’s account of principled negotiation distinguishes stated positions from underlying needs and concerns, and encourages parties to explore those interests.[1] The operative word is explore. A system can organize statements and suggest a question. It cannot recover a private reason simply by placing a demand inside an ontology.

Keep statements, interpretations, and unknowns separate

Suppose the operator later adds, “Daytime inspection would interrupt the refrigerated shipments.” That is evidence of a stated concern, attributed to the operator. It still does not prove the forecast disruption, establish that the concern is exhaustive, or rule out another motive. The same words can support one record while leaving another unresolved.

Our proposed representation separates direct statements from analyst hypotheses. Each hypothesis needs its supporting material, the person or process that proposed it, a review status, and alternatives worth considering. ODNI’s standards offer a relevant discipline: distinguish underlying information from assumptions and judgments, and explain which indicators would change an assessment.[2] We apply that principle as a design choice rather than a certification claim.

A graph makes the interpretation inspectable

The useful graph connection is not simply “operator has interest.” It is a reviewable assertion that a named statement supports a particular account of a concern. A separate relation can record that an analyst proposed an additional explanation. W3C PROV-O’s treatment of derivation and responsible agents is relevant to keeping that history.[3] These are proposed records, not a published TACITUS schema or a claim that our current implementation enforces them.

This structure would let a reviewer ask which interests were directly stated, which were inferred from the same passage, and which have no supporting source. It could also show that a supposed change of motive is really a different spokesperson discussing a different issue. A well-connected graph can still contain an incorrect interpretation; the links make it easier to examine, not more authoritative.

Overlap is a prompt for discussion, not a solution

A safety authority and a port operator may both express concern about keeping the port open. That common wording does not establish agreement about acceptable risk, inspection quality, or who bears the costs. Even a genuine shared interest can coexist with incompatible obligations. The next useful step is to ask about options and constraints, not to announce that the graph has found a resolution.

The UN mediation guidance emphasizes consent, impartiality, and the particular circumstances of each process.[4] Those requirements resist a purely computational account of agreement. In defensive narrative or statecraft analysis, the same caution applies to explanations of an institution’s behavior. A public demand, a policy objective, and an analyst’s account of strategic intent should remain distinguishable.

Evaluate the boundary between a clue and a conclusion

A research evaluation could present practitioners with fictional passages containing explicit concerns, ambiguous demands, sarcasm, reported speech, and deliberate gaps. Annotators should mark what the passage supports and record defensible disagreement. We would measure invented interests, incorrect attribution, and cases where an uncertain interpretation is presented as established fact. Abstaining on an unsupported motive is a valid outcome.

A second exercise could add a later statement that contradicts the original hypothesis and test whether the record and draft analysis change appropriately. In the proposed TACITUS stack, candidate interpretations would pass through DIALECTICA review before entering capsules used by the PRAXIS agentic workspace. The purpose is to support careful questions and revisable assessments. Evidence that the workflow improves those tasks still needs to be produced.

A documented request is not a private motive

S02 establishes that the Industry Forum requested a transition period. It does not prove why each member wanted one, that the request was accepted or that an entire sector shares one interest. Label any inferred interest as a hypothesis, name the supporting observations and alternative explanations, and ask what evidence could change it. The actor table separates proposer, consulted party, reviewer and decision owner without claiming access to hidden intent. [5]