EXP-0008 — Battery audit: twelve popular numeric claims under three fixed rules
completed Deterministic computation
Question
Which of the widely circulated numeric claims about the Qur'an hold when one set of counting rules is declared first and applied to every claim identically?
Hypothesis
No directional hypothesis. Popular claims are expected to split: some hold under a clean declared rule, most depend on bespoke counting choices (extends EXP-0002/0003).
Counting policy
Three rule kinds fixed for every claim before any count was run: letters-basic surface (bare + definite-article form only, no clitic variants), diacritized QAC lemma, QAC root. The lemma inventory was consulted during spec design solely to copy exact orthography. Empty rule lists (qul/qalu lemma+root) are declared not-applicable, recorded as zero, and cannot satisfy a claim.
Reproducibility record
- Corpus
- quran:hafs-kufan:v1
- Method version
- claims-battery-v1
- Results sha256
- 457a8b4b4d5808f6ff688bfc59134d2c…
- Environment
- {"platform":"Linux-6.18.5-fc-v20-x86_64-with-glibc2.39","python":"3.11.15"}
- Ran
- 2026-08-14T07:45:05Z → 2026-08-14T07:45:07Z
Re-run with python3 services/research-worker/scripts/run_experiment.py data/research/specs/EXP-0008.json. Input checksums are recorded in the result file.