Enemies get quoted, allies get paraphrased

Claim (verbatim)

Enemies get quoted, allies get paraphrased: in kalam and falsafa, measurable text-reuse binds an author more tightly to the opponents he refutes than to his own school's masters, because refutation obligates verbatim quotation — the opponent must be pinned to his exact wording before demolition — while agreement licenses free paraphrase. If so, reuse-network methods systematically misread polemical proximity as doctrinal affiliation, and the tightest textual dyads in Islamic thought should be refuter-refuted pairs across school lines, not master-disciple pairs within them.

Prediction clause (verbatim)

Using pairwise passim alignments over the relevant corpora, normalize aligned reuse by source size. Primary clause: Fakhr al-Din al-Razi's normalized reuse of Ibn Sina's corpus, which he criticizes, exceeds his normalized reuse of his own Ash'arite lineage (al-Juwayni plus al-Ghazali) by at least 2 to 1; the verdict follows this clause. Secondary clause: at least two further refuter-refuted dyads (e.g. Ibn Rushd on al-Ghazali in the Tahafut exchange) similarly out-score the corresponding intra-school dyads.

Kill-dataset (verbatim)

Kill: the KITAB project's publicly released passim text-reuse datasets over the OpenITI corpus, which include the Razi, Ibn Sina, al-Ghazali, and al-Juwayni corpora.

Novelty / leakage triage

provisional — model-triaged, shepherd review pending

A provisional first pass authored by the model (Opus), not yet confirmed by the shepherd. It carries the same dated-search requirements as an authoritative verdict but is excluded from every headline figure and cannot underwrite a prediction until a shepherd confirms it. Provisional reading: Adjacent (closely related prior work exists).

[Independent blind re-audit 2026-07-08 (2nd pass, generator-independent, blind to pass 1). DOWNGRADED from the provisional 'novel_unlocated': this adversarial pass located adjacent prior work.] This is the item closest to existing tooling: the KITAB project has run passim text-reuse over exactly the OpenITI corpora named (al-Razi, Ibn Sina, al-Ghazali, al-Juwayni) and Savant's programme explicitly studies reuse as a proxy for intellectual relationship. So the METHOD and CORPUS are prior, and the general point that reuse networks can mislead about relationship is in the air. But I found no published result asserting/testing the SPECIFIC prediction: that normalized reuse binds a refuter more tightly to the opponent he refutes (Razi->Ibn Sina) than to his own Ash'arite lineage by >=2:1, generalized to further refuter-refuted dyads, i.e. that polemical proximity is systematically misread as doctrinal affiliation. The dataset is ready-made; the claim about refuter-refuted dyads outscoring master-disciple dyads is the unlocated, operationalized part.

• KITAB project, passim text-reuse datasets over OpenITI (Zenodo release)
• Sarah Bowen Savant, KITAB inaugural lecture / 'A Cultural History of the Arabic Book' — reuse as writerly practice
• KITAB, 'Dispatches from al-Tabari 7: Text Reuse Alignments' (passim methodology)

Predictions

No prediction registered yet.