What does Computing Native Chinese Copy conclude, in numbers, about “Converting Between Malaysian, Singaporean, Taiwanese and Hong Kong Chinese: a Lexicon, an Affinity Matrix and a Library of Closing Sentences”?
Answer: Chapter 6 builds a small lexicon of confident pairs only for Malaysian, Singaporean, Taiwanese and Hong Kong Chinese; a mainland draft read in Taiwan has an illustrative probability of 0.968 of being noticed as written elsewhere.
Scope of the numbers: Its example sentences and parameters are illustrative; all calculations are generated by its programs; it is not individual legal, advertising-compliance or marketing advice, and it discusses only persuasion that keeps the rules. Chapter 2 proves that the dependency length of stacked modifiers grows roughly with the square of the layers, so that at four layers the illustrative probability of keeping every relation is only 0.098.
Data table
The numbers below are printed verbatim by the chapter's Python program and can be reproduced by running it. printed values that match the answer: 0.968
| m = foreign words / P(reads as written elsewhere), q = 0.35 | |||||
| copy \ reader | CN | TW | SG | MY | HK |
| CN | 0 / 0.000 | 8 / 0.968 | 4 / 0.821 | 4 / 0.821 | 5 / 0.884 |
| TW | 8 / 0.968 | 0 / 0.000 | 3 / 0.725 | 2 / 0.577 | 3 / 0.725 |
| SG | 4 / 0.821 | 6 / 0.925 | 0 / 0.000 | 1 / 0.350 | 3 / 0.725 |
| MY | 4 / 0.821 | 6 / 0.925 | 2 / 0.577 | 0 / 0.000 | 4 / 0.821 |
| HK | 5 / 0.884 | 6 / 0.925 | 3 / 0.725 | 3 / 0.725 | 0 / 0.000 |
| CN copy read in TW, q = 0.2: 0.832 | |||||
| CN copy read in TW, q = 0.5: 0.996 |
Show the full program
# Chapter 6. p62: An affinity matrix. Five versions of the same copy (the
# mainland draft and its four conversions by the lexicon of p61) are read
# by readers of five regions. m = the number of words in the copy that
# differ from the reader's own confident regional form (cells without a
# confident form are not counted). If a reader notices each foreign word
# independently with probability q, the chance that the copy reads as
# "written elsewhere" is 1 - (1 - q) ** m. q = 0.35 is illustrative; the
# pre-registered test of Chapter 8 is meant to estimate it.
LEX = { # concept: (CN, TW, SG, MY, HK)
"metro": ("地铁", "捷运", "地铁", "捷运", "港铁"),
"taxi": ("出租车", "计程车", "德士", "德士", "的士"),
"bicycle": ("自行车", "脚踏车", "脚车", "脚车", "单车"),
"bus": ("公交车", "公车", "巴士", "巴士", "巴士"),
"aircon": ("空调", "冷气", "冷气", "冷气", "冷气"),
"market": ("菜市场", "传统市场", "巴刹", "巴刹", "街市"),
"video": ("视频", "影片", None, None, None),
"software": ("软件", "软体", None, None, None),
"quality": ("质量", "品质", None, None, None),
"info": ("信息", "资讯", "资讯", None, "资讯"),
"potato": ("土豆", "马铃薯", None, None, "薯仔"),
}
REGIONS = ("CN", "TW", "SG", "MY", "HK")
Q = 0.35
def version(region):
"""Concept -> word used by the copy converted for this region."""
r = REGIONS.index(region)
used = {}
for concept in ("metro", "taxi", "bicycle", "aircon", "video",
"software", "quality", "info"):
forms = LEX[concept]
used[concept] = forms[r] if forms[r] is not None else forms[0]
return used
def mismatches(used, reader):
r = REGIONS.index(reader)
return len([c for c, w in used.items()
if LEX[c][r] is not None and LEX[c][r] != w])
print("m = foreign words / P(reads as written elsewhere), q = {}".format(Q))
print("copy \\ reader " + "".join("{:>11}".format(r) for r in REGIONS))
for src in REGIONS:
used = version(src)
cells = []
for reader in REGIONS:
m = mismatches(used, reader)
cells.append("{} / {:.3f}".format(m, 1 - (1 - Q) ** m))
print("{:<14}".format(src) + "".join("{:>11}".format(c) for c in cells))
for q in (0.2, 0.5):
m = mismatches(version("CN"), "TW")
print("CN copy read in TW, q = {}: {:.3f}".format(q, 1 - (1 - q) ** m))Single source and citation
Fan, Y. (2026). Native Cross-Context Prosody, De-Translationese Syntax-Tree Rewriting and a Conversion Corpus of Malaysian, Singaporean and Taiwanese Chinese (Version 1.0.0), Chapter 6: Converting Between Malaysian, Singaporean, Taiwanese and Hong Kong Chinese: a Lexicon, an Affinity Matrix and a Library of Closing Sentences. 5D Library. DOI pending
DOI pending · monograph page goes live after publication · 5DL-B-144 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026144,
author = {Fan, Yafu},
title = {Native Cross-Context Prosody, De-Translationese Syntax-Tree Rewriting and a Conversion Corpus of Malaysian, Singaporean and Taiwanese Chinese},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}Other chapter answers in this book
- Chapter 1 · Dependency Distance: Why Europeanised Long Sentences Are Hard to Read
- Chapter 2 · Chains of De and Stacked Modifiers: an Exact Account of the Left-Branching Burden
- Chapter 3 · Fingerprints of Translationese: Ten Machine-Checkable Markers and a Poisson Likelihood-Ratio Test
- Chapter 4 · Rewriting the Tree: Three Operations and the Long-Outside Ordering Theorem
- Chapter 5 · Prosody and Breath: Segment Lengths, Dynamic Programming for Commas, and the Rhythm of Fours and Sixes
- Chapter 7 · The Classics Examined: Sentences and Pruning in the Wenxin Diaolong, and Orwell's Six Rules
- Chapter 8 · Practice and Refutation: the Persuasive-Copy Checker, a Ceiling on Private Follow-Ups, and a Pre-Registered Test