| paper | d4-conf-ijcai-1989-0070 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | unclear |
The paper contains no named computational result to anchor a mirror to. A continuous computational study of opinion aggregation could be devised, but its procedure and question would need to be specified beyond what this paper provides.
fails bit a — no named computational result to mirror
The paper describes the COMBINE, DOUBT, and REINTERPRET modules and gives an illustrative example, but states no named computational result.
The strongest honest positive case is only a plausible research direction, not a mirror anchored in this paper’s results. The paper contains no numbered theorem, lemma, corollary, or proposition asserting a computational complexity result. Sections 3.1–3.4 describe the COMBINE, DOUBT, and REINTERPRET procedures; Section 4 gives an illustrative example with arbitrary attribute values. None supplies a complexity anchor, so there is no result here for which I can honestly state a continuous problem and predict its place in the tractable / hardness-transfers / continuum-specific-hard trichotomy.
There is a plausible population regime behind the model: a large public consultation or administrative evidence pool addressing one proposition. A type could comprise a source’s opinion certainty, topic-specific competence, power, goals, and relevant scripts; \(\mu_t\) would be the fraction of reports with that complete profile. The continuous task would process those source-type masses through credibility adjustment and opinion combination, then return a certainty for the proposition and the resulting judgment. That is recognizably related to the paper’s aim of forming a fact from credibility-weighted opinions.
The weakness is decisive for this assignment: the paper does not specify the combination and update rules precisely enough to make that task a prover-ready computational problem. It also does not establish a computational result to mirror. Turning the conceptual procedure into a formal continuous problem could be worthwhile, but presenting it as a mirror of a named result here would overstate the case.
There are no named results to anchor a continuous mirror to. The paper states no computational problem, complexity bound, or algorithmic guarantee. It describes a cognitive architecture—COMBINE, DOUBT, and REINTERPRET—with heuristic choices, configurable modes, and an illustrative example whose values are arbitrary. That makes the proponent’s admission decisive for this assignment: the paper offers no computational result for the programme to mirror.
The proposed large consultation is a plausible high-multiplicity setting. One could define types by opinion certainty, competence, power, goals, and scripts, then ask how a distribution of reports affects a judgment. But even this stronger version needs choices the paper does not settle: how opinions combine, how credibility changes, how doubt and reinterpretation are selected, and how the cycle stops. A continuous distribution can be passed through any specified procedure; the resulting complexity question would belong to the new procedure an analyst defines, not to a result in this paper.
One could also add an intervention question, such as how much evidence must change to flip the judgment. That would create a computational problem closer to bribery or control, but it would be a new research model. The paper does not define such an intervention or claim a result about it.
So the negative case is strong about *mirroring this paper*, but it cannot establish that no worthwhile continuous computational study of opinion aggregation is possible. The consultation scenario could motivate one. It simply would not be a continuous mirror of a computational result that the paper contains.
The adversarial triple: the proponent anchors on up to three named results; the opponent sees that case and must defeat every anchor; the judge decides which case convinced it. These are the pipeline’s own outputs, generated by tools/triple_run.py — no human edited them. The paper’s own text is not reproduced here beyond the quoted statement above.