关于「原典考辨:咸池三闻、声无哀乐与汉斯立克的「乐音运动的形式」」,《和声惊奇、谢帕德音与音乐战栗》给出的定量结论是什么?
结论:咸池三闻的「惧、怠、惑」,在学习型听众中对应为:起初惊奇度 2.315 比特;熟悉后降到 2.042 比特、超额惊奇只有 0.086;信源改变后升到 3.167 比特、超额惊奇 1.085。嵇康的「心之與聲,明為二物」在信息论中成立:同一段音乐对两位听众的惊奇度相关系数只有 −0.10;但响度对两人都是 32 宋,唤醒确实在声。汉斯立克把音乐的内容限定为乐音运动的形式,与嵇康的「躁静」惊人地相近,而两人都低估了期待本身的力量。
数据口径:书中参数为示意;全部计算由书中程序生成;本书不是医疗建议,不宣称任何健康效果,也不是营销合规或法律意见。 第一章把和声写成马尔可夫链:在示意矩阵中,阻碍终止带来 3.06 比特的惊奇,正格终止只有 0.51 比特。
对照数据表
下表数字由本章 Python 程序原样打印,可重新运行复现。 程序只用 Python 标准库与结论吻合的打印值: 0.086, 1.085, 2.042, 2.315, 3.167
| block chords phase surprisal pred. entropy excess | |
| 1 | 1-40 fear 2.315 2.249 0.065 |
| 2 | 41-80 slack 2.073 2.029 0.045 |
| 3 | 81-120 slack 2.053 1.969 0.084 |
| 4 121-160 slack | 2.042 1.956 0.086 |
| 5 161-200 bewilder | 3.167 2.082 1.085 |
| 6 201-240 bewilder | 2.989 2.207 0.782 |
| true entropy rate of the tonal source 1.913; uniform 2.585 |
查看完整验算程序
# Chapter 7. p71: "first fear, then slackness, at last bewilderment"
# (Zhuangzi, Tian yun) as a learning listener. The listener predicts
# each chord from Dirichlet counts (prior 0.5 per cell, six chords).
# Chords 1 to 160 come from the ILLUSTRATIVE tonal matrix of p11;
# chords 161 to 240 from a source with no tendencies (uniform), the
# "no tail at the end, no head at the start" of the text.
# Per block of 40: mean surprisal, mean predictive entropy of the
# listener, and their gap (excess surprise), averaged over 200
# simulated listeners with a fixed seed.
import math
import random
CH = ["I", "ii", "IV", "V", "vi", "bVI"]
P = {
"I": [0.05, 0.15, 0.30, 0.30, 0.15, 0.05],
"ii": [0.05, 0.05, 0.05, 0.75, 0.05, 0.05],
"IV": [0.25, 0.10, 0.05, 0.45, 0.10, 0.05],
"V": [0.70, 0.02, 0.05, 0.08, 0.12, 0.03],
"vi": [0.05, 0.30, 0.35, 0.20, 0.05, 0.05],
"bVI": [0.20, 0.05, 0.25, 0.40, 0.05, 0.05],
}
UNI = [1 / 6] * 6
RUNS = 200
rng = random.Random(20261112)
acc = [[0.0, 0.0] for _ in range(240)]
for _ in range(RUNS):
cnt = {c: [0.5] * 6 for c in CH}
prev = "I"
for n in range(240):
src = P[prev] if n < 160 else UNI
nxt = rng.choices(CH, weights=src)[0]
tot = math.fsum(cnt[prev])
pred = [x / tot for x in cnt[prev]]
acc[n][0] += -math.log2(pred[CH.index(nxt)]) / RUNS
acc[n][1] += -math.fsum(q * math.log2(q) for q in pred) / RUNS
cnt[prev][CH.index(nxt)] += 1
prev = nxt
rows = acc
NAMES = ["fear", "slack", "slack", "slack", "bewilder", "bewilder"]
print("block chords phase surprisal pred. entropy excess")
for b in range(6):
blk = rows[40 * b:40 * b + 40]
s = math.fsum(r[0] for r in blk) / 40
h = math.fsum(r[1] for r in blk) / 40
print("{:>5} {:>3}-{:<4} {:<10} {:>9.3f} {:>14.3f} {:>7.3f}".format(
b + 1, 40 * b + 1, 40 * b + 40, NAMES[b], s, h, s - h))
print("true entropy rate of the tonal source 1.913; uniform {:.3f}".format(
math.log2(6)))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《管弦配乐和声期待违背熵、谢帕德无限音阶错觉与音乐战栗立毛反射动力学》第 7 章「原典考辨:咸池三闻、声无哀乐与汉斯立克的「乐音运动的形式」」(版本 1.0.0). 5D Library. DOI 待发布
DOI 待发布 · 专著页正式发布后上线(本站暂无专著页) · 5DL-B-163 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026163,
author = {Fan, Yafu},
title = {Harmonic-Expectation Violation Entropy in Orchestral Scoring, the Shepard-Tone Illusion and the Dynamics of Musical Chills},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}