What does Library Conflicts, Reproducibility and Erasure Codes conclude, in numbers, about “Classical Sources: Siku Collation, the Fate of the Seven Pavilions and Lachmann's Stemmatics”?
Answer: Chapter 7 examines the collation records of the Siku quanshu and the fate of its seven pavilion copies alongside Lachmann's stemmatics: with four kindred copies and one independent copy, voting by manuscripts leaves 20.300 wrong characters per thousand and comparing families only 0.100.
Scope of the numbers: Library sizes, error rates, hazards and failure rates are illustrative; all calculations are generated by the programs in the book; the book is not legal, financial, procurement or storage-service advice. Chapter 1 draws the library as a knowledge graph and shows that the more volumes there are, the fewer places an error can hide: at 200 volumes, 0.9821 of deviant definitions in the illustrative library sit where another volume can contradict them.
Data table
The numbers below are printed verbatim by the chapter's Python program and can be reproduced by running it. standard-library Python onlyprinted values that match the answer: 0.100, 20.300
| reading | wrong / 1000 | flagged/1000 |
| one copy | 20.800 | 0.000 |
| 3 independent copies, plurality | 0.000 | 1.300 |
| 5 independent copies, plurality | 0.000 | 0.050 |
| 7 independent copies, plurality | 0.000 | 0.000 |
| 4 from beta + 1 alpha, naive plurality | 20.300 | 0.100 |
| 4 from beta + 1 alpha, Lachmann two branches | 0.100 | 39.800 |
| code view: 3 copies = repetition code, rate 0.333, fixes 1 error; | ||
| RS(6+3) has rate 0.667, fixes 3 erasures or 1 error per stripe |
Show the full program
# Chapter 7. p72: collation as decoding. A text of 20,000 characters is
# copied by hand; each copy miswrites a character with probability E.
# A miswriting is the common slip (one shared wrong reading, such as a
# look-alike character) with probability SHARE, otherwise a private one.
# (A) independent witnesses copied from the archetype, read by plurality;
# (B) five witnesses of which four descend from one lost exemplar (beta)
# and one from the archetype (alpha): naive plurality against Lachmann's
# rule (rebuild beta from its four, then compare beta with alpha, and
# flag every disagreement for the editor). E and SHARE are illustrative.
import random
from collections import Counter
rng = random.Random(20261112)
L, E, SHARE = 20000, 0.02, 0.3
def copy(text):
out = []
for ch in text:
if rng.random() < E:
if rng.random() < SHARE:
out.append(-ch - 1)
else:
out.append(rng.randrange(10 ** 6, 10 ** 7))
else:
out.append(ch)
return out
def plural(wits):
res = []
for col in zip(*wits):
c = Counter(col).most_common()
if len(c) > 1 and c[0][1] == c[1][1]:
res.append(None)
else:
res.append(c[0][0])
return res
def score(read, truth):
wrong = len([1 for r, t in zip(read, truth) if r is not None and r != t])
flag = len([1 for r in read if r is None])
return 1000 * wrong / L, 1000 * flag / L
arch = list(range(L))
print("{:<44} {:>13} {:>13}".format(
"reading", "wrong / 1000", "flagged/1000"))
print("{:<44} {:>13.3f} {:>13.3f}".format(
"one copy", *score(copy(arch), arch)))
for n in (3, 5, 7):
wits = [copy(arch) for _ in range(n)]
print("{:<44} {:>13.3f} {:>13.3f}".format(
"{} independent copies, plurality".format(n),
*score(plural(wits), arch)))
beta = copy(arch)
fam = [copy(beta) for _ in range(4)] + [copy(arch)]
print("{:<44} {:>13.3f} {:>13.3f}".format(
"4 from beta + 1 alpha, naive plurality", *score(plural(fam), arch)))
b_hat = plural(fam[:4])
lach = [b if b == a else None for b, a in zip(b_hat, fam[4])]
print("{:<44} {:>13.3f} {:>13.3f}".format(
"4 from beta + 1 alpha, Lachmann two branches", *score(lach, arch)))
print("code view: 3 copies = repetition code, rate 0.333, fixes 1 error;")
print("RS(6+3) has rate 0.667, fixes 3 erasures or 1 error per stripe")Single source and citation
Fan, Y. (2026). Cross-Volume Theorem-Conflict Detection in Large Bilingual Monograph Libraries, Anti-Decay of Code Reproducibility and Erasure-Coded Cold Storage (Version 1.0.0), Chapter 7: Classical Sources: Siku Collation, the Fate of the Seven Pavilions and Lachmann's Stemmatics. 5D Library. DOI pending
DOI pending · monograph page goes live after publication · 5DL-B-182 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026182,
author = {Fan, Yafu},
title = {Cross-Volume Theorem-Conflict Detection in Large Bilingual Monograph Libraries, Anti-Decay of Code Reproducibility and Erasure-Coded Cold Storage},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}Other chapter answers in this book
- Chapter 1 · The Library as a Graph: Volumes, Chapters, Terms and Numeric Claims as Nodes
- Chapter 2 · Normalization and Hash Indexing: False Alarms and Misses in Cross-Volume Conflict Detection
- Chapter 3 · The Sampling Theory of Regression Audits: the Rule of Three and the Finite-Population Correction
- Chapter 4 · The Decay of Code Reproducibility: Hazard Rates of Dependency Drift and Byte-Exact Output Regression
- Chapter 5 · Reed-Solomon over GF(2^8): Erasure Coding and Recovery in the Standard Library
- Chapter 6 · The Reliability of Replicas and Erasure Codes: Binomial Windows, Markov Chains and Scrubbing for Bit Rot
- Chapter 8 · Practice and Falsification: a Publisher's Checklists for Consistency, Reproducibility and Cold Storage