What does Lexical Coverage and the Limits of Fast Tracks conclude, in numbers, about “Closing the Loop: Customer-Service Word Lists, Team Learning Plans and How to Refute This Book”?

Answer: Chapter 8 gives a procedure for building, measuring and re-measuring word lists for service teams in Malaysia and Taiwan, the distilled Word-List Coverage Calculator (in its demonstration the affix rules raise coverage from 0.6667 to 0.8667), three pre-registered tests and the results that would refute the book's central claim. Marked boxes separate published evidence, runnable Python calculations (19 programs), classical sources and the author's hypotheses.

Scope of the numbers: Its parameters are illustrative and all calculations are generated by its programs. It is general reference material for owners and teachers, not business advice for any person or firm, and it promises no fluency from any course within any period.

中文版

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: 0.6667, 0.8667

Malay, affix rules off: tokens 30, known 20, coverage 0.6667
  unknown: dibuat(1) dihantar(1) menghubungi(1) menunggu(1)
    pembayaran(1) penghantaran(1) pesanan(1) sebentar(1) talian(1)
    termasuk(1)
  unknown per 300-word page: 100.0
  list multiplier for 0.95: x 5.7
  list multiplier for 0.98: x 10.2
Malay, affix rules on: tokens 30, known 26, coverage 0.8667
  unknown: menghubungi(1) sebentar(1) talian(1) termasuk(1)
  unknown per 300-word page: 40.0
  list multiplier for 0.95: x 2.2
  list multiplier for 0.98: x 3.9
Hokkien (Tai-lo, tone numbers): tokens 10, known 8, coverage 0.8000
  unknown: gua7-tse7(1) phainn2-se3(1)
  unknown per 300-word page: 60.0
  list multiplier for 0.95: x 3.2
  list multiplier for 0.98: x 5.7
Show the full program
# Chapter 8. p82: the core of the word-list coverage calculator
# Tokens are runs of letters and digits, hyphenated runs kept whole, so
# Tai-lo with tone numbers (to1-sia7) counts as one word. A token is
# known if it is on the list or, in Malay mode, if the affix rules of
# p32 reduce it to a root on the list. Coverage = known / tokens. The
# projection finds the list size n_eq on the reference curve of p23
# (V = 20000, s = 2.389, q = 664.5) that gives the measured coverage
# and scales the list by n_target / n_eq; it is a rough planning
# figure, not a measurement. Demo texts and lists are hand-made.
import math
import re
import textwrap
from collections import Counter

V, S, Q = 20000, 2.389, 664.5
PREFIX = [("meny", "s"), ("meng", ""), ("meng", "k"), ("mem", ""),
          ("mem", "p"), ("men", ""), ("men", "t"), ("me", ""),
          ("peny", "s"), ("peng", ""), ("peng", "k"), ("pem", ""),
          ("pem", "p"), ("pen", ""), ("pen", "t"), ("pel", ""),
          ("per", ""), ("pe", ""), ("bel", ""), ("ber", ""), ("di", ""),
          ("ter", ""), ("", "")]
SUFFIX = ["kan", "an", "i", ""]


def tokens(text):
    return re.findall(r"[a-z0-9]+(?:-[a-z0-9]+)*", text.lower())


def malay_root_known(word, known):
    for pre, restore in PREFIX:
        if word.startswith(pre):
            stem = restore + word[len(pre):]
            for suf in SUFFIX:
                if suf == "" or stem.endswith(suf):
                    root = stem[:len(stem) - len(suf)] if suf else stem
                    if root in known:
                        return True
    return False


def coverage(text, wordlist, malay=False):
    known = set(tokens(wordlist))
    toks = tokens(text)
    miss = Counter()
    for t in toks:
        if t in known or (malay and malay_root_known(t, known)):
            continue
        miss[t] += 1
    return len(toks), len(toks) - sum_counts(miss), miss


def sum_counts(counter):
    n = 0
    for v in counter.values():
        n += v
    return n


CURVE = [(r + Q) ** (-S) for r in range(1, V + 1)]
TOTAL = math.fsum(CURVE)


def n_for(c):
    lo, hi = 0, V
    while hi - lo > 1:
        mid = (lo + hi) // 2
        if math.fsum(CURVE[:mid]) / TOTAL >= c:
            hi = mid
        else:
            lo = mid
    return hi


def report(name, text, wordlist, malay):
    n, k, miss = coverage(text, wordlist, malay)
    c = k / n
    print("{}: tokens {}, known {}, coverage {:.4f}".format(name, n, k, c))
    words = " ".join("{}({})".format(w, miss[w]) for w in sorted(miss))
    print(textwrap.fill("unknown: " + words, width=72,
                        initial_indent="  ", subsequent_indent="    "))
    print("  unknown per 300-word page: {:.1f}".format(300 * (1 - c)))
    neq = n_for(c)
    for tgt in (0.95, 0.98):
        print("  list multiplier for {:.2f}: x {:.1f}".format(
            tgt, n_for(tgt) / neq))


MS_TEXT = ("Selamat pagi, boleh saya bantu? Terima kasih kerana menunggu."
           " Pesanan anda sudah dihantar. Harga ini termasuk penghantaran."
           " Sila tunggu sebentar. Kami akan menghubungi anda esok."
           " Pembayaran boleh dibuat dalam talian.")
MS_LIST = ("selamat pagi boleh saya bantu terima kasih kerana tunggu "
           "pesan anda sudah hantar harga ini sila kami akan esok bayar "
           "buat dalam")
HOK_TEXT = ("li2 ho2. to1-sia7. phainn2-se3. tsit4 e5 gua7-tse7 tsinn5?"
            " u7 bo5?")
HOK_LIST = "li2 ho2 to1-sia7 tsit4 e5 tsinn5 u7 bo5"

if __name__ == "__main__":
    report("Malay, affix rules off", MS_TEXT, MS_LIST, False)
    report("Malay, affix rules on", MS_TEXT, MS_LIST, True)
    report("Hokkien (Tai-lo, tone numbers)", HOK_TEXT, HOK_LIST, False)

Single source and citation

Fan, Y. (2026). The 98% Lexical-Coverage Threshold, Zipfian Frequency Curves and Fast-Track Limits for Malay and Hokkien Learners (Version 1.0.0), Chapter 8: Closing the Loop: Customer-Service Word Lists, Team Learning Plans and How to Refute This Book. 5D Library. DOI pending

DOI pending · monograph page goes live after publication · 5DL-B-228 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026228,
  author    = {Fan, Yafu},
  title     = {The 98% Lexical-Coverage Threshold, Zipfian Frequency Curves and Fast-Track Limits for Malay and Hokkien Learners},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

Other chapter answers in this book