5D Library › 多阶段最后通牒议价、应收账款衰减动力学与最优施压序列 · English

第七章 烛之武、《谋攻》与《唐律》的负债阶梯

为什么在这里读这些史料

前几章有三个主题:改变对方对其自身替代选择的估计;逐步施压、只施加到必要为止;以及随债务逾期时间而加重的处罚。中国史料对三者都有记载。本章引出原文,看哪些对应可以计算。核对后要说明两点:常被说成出自《战国策》的烛之武说秦伯,实见于《左传·僖公三十年》;唐宋的「公验」是官府发给的凭证(例如通行凭证),不是债务追偿制度,唐代的债务规定见于《唐律疏议·杂律》。

烛之武:重新框定对方的替代选择

原典

《左传·僖公三十年》

秦、晉圍鄭,鄭既知亡矣。若亡鄭而有益於君,敢以煩執事。越國以鄙遠,君知其難也。焉用亡鄭以陪鄰?鄰之厚,君之薄也。若舍鄭以為東道主,行李之往來,共其乏困,君亦無所害。

(杨伯峻 1981)

烛之武既没有威胁秦伯,也没有贿赂他;他改变的是秦伯对每个选择价值的估计。p13 的 (b) 部分给它配上数字。

推导

# Chapter 7. p13: Three old ladders as arithmetic.
# (a) Tang code, "Za lu": a debt unpaid 20 days past the contract gets
# 20 strokes of the light stick; each further 20 days adds one grade,
# up to 60 strokes of the heavy stick; 30 bolts or more add two grades,
# 100 bolts or more three grades more; and the debt must still be
# repaid. Grades: light stick 10 to 50, heavy stick 60 to 100, then
# penal servitude of 1 year, 1.5 years and so on.
# (b) Zhu Zhiwu to the Duke of Qin: Qin and Jin each have strength 10
# and Zheng 4. If Zheng is destroyed, Jin takes its land (strength 14)
# because Qin cannot hold land beyond Jin; if Zheng is spared, it
# serves Qin as a host on the eastern road (worth 1 a year to Qin).
# Qin's strength relative to Jin under each choice.
# (c) Sunzi, "Mou gong": the four levels of attack with illustrative
# costs 1, 3, 10, 30 and chances of success 0.3, 0.4, 0.6, 0.8. Trying
# the levels in order, stopping at the first success.
GRADES = ["light 10", "light 20", "light 30", "light 40", "light 50",
          "heavy 60", "heavy 70", "heavy 80", "heavy 90", "heavy 100",
          "1 yr", "1.5 yrs"]

def tang(days, bolts):
    if days < 20:
        return "none"
    g = min(1 + (days - 20) // 20, 5)       # base: 20 strokes .. 60
    if bolts >= 30:
        g += 2
    if bolts >= 100:
        g += 3
    return GRADES[g]

if __name__ == "__main__":
    print("   (a) days late    1 bolt     30 bolts    100 bolts")
    for d in (20, 40, 60, 80, 100, 140):
        print("   {:13d}    {:<9}  {:<10}  {}".format(
            d, tang(d, 1), tang(d, 30), tang(d, 100)))
    print("   (b) Qin / Jin if Zheng destroyed: {:.3f}".format(10 / 14))
    print("       Qin / Jin if Zheng spared (5 years): {:.3f}".format(
        (10 + 5) / 10))
    costs, wins = (1, 3, 10, 30), (0.3, 0.4, 0.6, 0.8)
    left, exp_cost = 1.0, 0.0
    for c, w in zip(costs, wins):
        exp_cost += left * c
        left *= 1 - w
    print("   (c) in order: expected cost {:.2f}, failure {:.3f}".format(
        exp_cost, left))
    print("       straight to the walls: cost 30, failure {:.3f}".format(
        0.2))
#    (a) days late    1 bolt     30 bolts    100 bolts
#               20    light 20   light 40    heavy 70
#               40    light 30   light 50    heavy 80
#               60    light 40   heavy 60    heavy 90
#               80    light 50   heavy 70    heavy 100
#              100    heavy 60   heavy 80    1 yr
#              140    heavy 60   heavy 80    1 yr
#    (b) Qin / Jin if Zheng destroyed: 0.714
#        Qin / Jin if Zheng spared (5 years): 1.500
#    (c) in order: expected cost 12.34, failure 0.034
#        straight to the walls: cost 30, failure 0.200

如果秦、晋各有实力 10,晋吞并郑的 4,秦相对于晋的实力降到 0.714;如果保全郑国并让它供应秦国,在这个示意中五年后比值是 1.500。秦伯撤兵了。这是第六章那一步的反面:不是改善自己的替代选择,而是让对方看到它的替代选择比它以为的更差。

孙子:升级的阶梯

原典

《孙子·谋攻》

是故百戰百勝,非善之善者也;不戰而屈人之兵,善之善者也。故上兵伐謀,其次伐交,其次伐兵,其下攻城。

(杨丙安 1999)

p13 的 (c) 部分给四个层次示意的成本 1、3、10、30,以及逐级上升的成功机会。依次尝试、第一次成功就停止,平均成本 12.34,失败的机会 0.034;直接攻城,成本 30,失败的机会 0.200。第四章的阶梯有同样的形状:从便宜而且保全关系的手段开始,只爬到必要的高度。

唐律:法律中的阶梯

原典

《唐律疏议·杂律》

诸负债违契不偿,一匹以上,违二十日笞二十,二十日加一等,罪止杖六十;三十匹加二等,百匹又加三等。各令备偿。

(长孙无忌 and 刘俊文 1983)

p13 的 (a) 部分列出这个阶梯。小额债务在第 100 天达到杖六十的上限;三十匹的债务起点高两等,上限也高两等;一百匹的债务从杖七十升到徒一年。处罚随时间与金额两者加重,而且从不取代偿还:这正是本书为现代升级顺序所建议的两个特点,当然要在今天的法律之内。

假说 · 尚未验证

作者的诠释(尚未验证): 唐律负债不偿的刑罚阶梯,是为了在私人执行不划算的小额债务上让债权人的威胁可信,正如今天的小额诉讼程序。什么会证明它错: 在存世的唐代与敦煌契约文书中,小额债务并不比大额债务更依赖官府处罚,或者契约改而依赖私人担保。

小结

烛之武靠改变秦伯对自身替代选择的看法取胜;《孙子》把便宜而间接的手段排在昂贵而直接的手段之前;唐律随时间与金额加重处罚,同时始终责令偿还。两种常见的说法,出自《战国策》与「公验」债务制度,都不成立。

参考文献

杨丙安. 1999. 十一家注孙子校理. 中华书局.
杨伯峻. 1981. 春秋左传注. 中华书局.
长孙无忌, and 刘俊文. 1983. 唐律疏议. 中华书局.

Fan, Y. (2026). Multistage Ultimatum Bargaining, Accounts-Receivable Decay Kinetics and Optimal Escalation Sequences: Ackerman Concession Curves, Credible Threat Thresholds and Default Recovery Bounds [多阶段最后通牒议价、应收账款衰减动力学与最优施压序列] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.24090245 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).