Rayna · business models & économie numérique

Pruning log — librarian / citation-arbiter

Append-only audit trail of references considered but not promoted to the corpus. Each entry names the ref, its three Promote-predicate values, and a one-sentence rationale. Promote(R) ⟺ |body_cites(R)| ≥ 2 ∧ flagged(R) ∧ topical(R).


Round 1 — task-20260620-60ab (2026-06-20)

Corpus at decision time: teece2010, chesbroughRosenbloom2002, rochetTirole2003 (reader articles) + rochetTirole2006 (PDF materialised, no reader). 18 distinct refs parsed from cites: frontmatter across 3 articles. 0 promoted, 18 pruned.

Two structural facts make the frontier empty before any per-ref test: 1. No ref is cited by ≥ 2 articles — every ref appears in exactly one article’s cites: set, so |body_cites(R)| ≤ 1 for all R. 2. No <!-- librarian:flag --> markers exist in any article body, so flagged(R) is false for every R.

Per-ref entries follow (sorted by ref):

R = andrews1987

R = armstrong2006

R = bayeMorgan2001

R = caillaudJullien2003

R = chandler1962

R = chesbroughRosenbloom2002 (already in corpus — PDF present)

R = christensen1997

R = farrellSaloner1985

R = katzShapiro1985

R = laffontEtAl2001

R = parkerVanAlstyne2000

R = prahaladBettis1986

R = rochetTirole2002

R = rochetTirole2006 (already in corpus — PDF present)

R = rohlfs1974

R = schmalensee2002

R = teece1986

R = weick1993