5D Library › Evolutionary Game Theory, Direct and Indirect Reciprocity and the Emergence of Cooperation · 中文
Chapter 7 Reading the Classics: Graded Love, Impartial Care and Repaying Resentment with Uprightness
This chapter reads the research of the earlier chapters beside several pre-Qin texts: the Mencius on “love one’s kin, and so be humane to the people; be humane to the people, and so cherish creatures”, together with Mencius’s criticism of Mohist “impartial care”; the Mozi, “Universal Love II”, on “whoever loves others will surely be loved by others in return”; and the Analects, “Xian Wen”, on “repay resentment with uprightness, repay kindness with kindness”. The aim is not to show that the ancients already understood Hamilton’s rule or evolutionary games, but to see how ancient thinkers discussed whether love should be graded, whether treating others well is repaid, and how to treat those who wrong us, and to state three boundaries. First, these texts are ethical and political claims about how people ought to act, whereas evolutionary game theory asks which strategies will be retained under selection; the questions differ. Second, the Confucian “love of kin” is a ritual ethic centred on the patrilineal family, while Hamilton’s coefficient of relatedness is a genetic statistic; the two resemble each other only in the form “give more to those who are closer”. Third, Mozi’s “whoever loves others will surely be loved in return” is persuasion, and its “surely” is rhetorical; modern research has found precisely that returns are conditional, not certain. All translations in this chapter are the author’s own.
Love one’s kin, and so be humane to the people
Source text
孟子曰:「君子之于物也,爱之而弗仁;于民也,仁之而弗亲。亲亲而仁民,仁民而爱物。」(杨伯峻 译注 1960)
(Mencius, “Jin Xin I”)
夷子曰:「儒者之道,古之人『若保赤子』,此言何谓也?之则以为爱无差等,施由亲始。」……孟子曰:「夫夷子信以为人之亲其兄之子为若亲其邻之赤子乎?……且天之生物也,使之一本,而夷子二本故也。」(杨伯峻 译注 1960)
(Mencius, “Teng Wen Gong I”; the middle of the passage is omitted)
杨氏为我,是无君也;墨氏兼爱,是无父也。无父无君,是禽兽也。(杨伯峻 译注 1960)
(Mencius, “Teng Wen Gong II”)
In the author’s translation: Mencius said: “Towards creatures the gentleman is sparing, but does not treat them with humaneness; towards the people he is humane, but does not love them as kin. He loves his kin, and so is humane to the people; he is humane to the people, and so cherishes creatures.” The Mohist Yi Zhi said: “The way of the Confucians is that the ancients cared for the people ‘as if protecting an infant’. What does this saying mean? I take it to mean that love has no gradations, but its practice begins with one’s kin.” … Mencius said: “Does Master Yi truly believe that a man’s love for his brother’s son is like his love for a neighbour’s infant? … Moreover, when Heaven gives birth to things it gives them a single root, but Master Yi has two roots.” Yang Zhu’s “each for himself” means having no ruler; Mozi’s “impartial care” means having no father. To have no father and no ruler is to be a beast.
Original meaning and context. The passage in “Jin Xin I” is about the order of love: the three words “love as kin”, “be humane” and “cherish” apply to kin, the people and creatures, in decreasing degree, but all extend outwards from the same heart. This is what later came to be called “love with gradations”. Yi Zhi was a Mohist who tried to reconcile the two schools: love itself has no gradations, but its practice begins with one’s kin. Mencius rejected the compromise: for him the gradations of love are not a practical convenience but come from human nature; a person has a single root (his parents), and to treat a neighbour’s child and a brother’s child alike is to have “two roots”. The criticism of the Mohists in “Teng Wen Gong II” is harsher still; it is the polemical language of Mencius’s debates, not an objective description of Mohist doctrine.
Read beside modern research. “Love one’s kin, and so be humane to the people” resembles Hamilton’s rule in one formal respect: one gives more to those who are close and less to those who are distant. In , an act of help costing is favoured by selection only if the recipient’s relatedness is high enough; the more distant the relative, the higher the required. But the differences matter more than the resemblance. First, Mencius’s gradations are a moral and ritual order that asks a person to “extend kindness”, carrying love for kin outwards to the people and to creatures; Hamilton’s rule describes changes in gene frequencies and carries no “ought to extend”. Second, Mencius takes the parents as the “single root”, the standpoint of a patrilineal ethic; relatedness is a symmetric statistic, the same (1/2) for siblings, children and parents, and gives parents no special place. Third, Chapter 2 showed that in Hamilton’s rule in regression form can come from any process that brings similar individuals together, not necessarily kinship. So it is wrong to say that “Mencius had already discovered kin selection”; what can be said is that Confucian ethics and kin-selection theory both noticed the fact that “how much one gives varies with the closeness of the relation”, and explained it in completely different ways.
Whoever loves others will surely be loved in return
Source text
子墨子言曰:「以兼相爱交相利之法易之。」……子墨子言:「视人之国,若视其国;视人之家,若视其家;视人之身,若视其身。」……夫爱人者,人必从而爱之;利人者,人必从而利之;恶人者,人必从而恶之;害人者,人必从而害之。(孙诒让 撰,孙启治 点校 2001)
(Mozi, “Universal Love II”; parts of the passage are omitted)
无言不雠,无德不报。……投我以桃,报之以李。(程俊英、蒋见元 注析 1991)
(Book of Songs, “Major Odes”, “Yi”)
姑尝本原之先王之所书,《大雅》之所道曰:「无言而不雠,无德而不报。」「投我以桃,报之以李。」即此言爱人者必见爱也,而恶人者必见恶也。(孙诒让 撰,孙启治 点校 2001)
(Mozi, “Universal Love III”)
In the author’s translation: Master Mozi said: “Replace it (mutual hatred and mutual harm) with the way of impartial mutual love and the exchange of mutual benefit.” … Master Mozi said: “Regard other people’s states as you regard your own state; regard other people’s houses as you regard your own house; regard other people’s persons as you regard your own person.” … Whoever loves others will surely be loved by others in return; whoever benefits others will surely be benefited by others in return; whoever hates others will surely be hated by others in return; whoever harms others will surely be harmed by others in return. The ode “Yi” says: no word goes unanswered, no kindness goes unrepaid … You throw me a peach, and I repay you with a plum. “Universal Love III” says: let us trace the matter to its source in the writings of the former kings; the “Major Odes” say, “No word goes unanswered, no kindness goes unrepaid”, “You throw me a peach, and I repay you with a plum”; this means that whoever loves others will surely be loved, and whoever hates others will surely be hated.
Original meaning and context. The three chapters on “Universal Love” are Mozi’s political programme against a Warring States world of rulers attacking one another and houses usurping one another. He held that all the disorder in the world arises from people “not loving one another”, and so advocated “impartial mutual love and the exchange of mutual benefit”. The passage “whoever loves others will surely be loved in return” answers contemporaries who objected that impartial care was “hard to practise”: Mozi replies that it is not hard, because whoever loves others will surely be loved by them, and whoever benefits others will surely be benefited. The argument has two levels: a moral claim (one ought to care impartially) and a claim about consequences (impartial care will be repaid). The second gives Mohist impartial care a clearly utilitarian colouring, which is one reason Mencius criticised it. “Universal Love III” also cites these lines of the ode “Yi” as support (quoted above) and concludes from them that “whoever loves others will surely be loved”: the ode speaks of words and kindness being answered, and Mozi turns it into evidence that impartial care is always repaid.
Read beside modern research. Mozi’s second argument has a structural parallel with the direct and indirect reciprocity of Chapters 3 and 4: “whoever loves others will surely be loved in return” is close to “treating others well is repaid”, and “whoever hates others will surely be hated in return” is close to the retaliation of tit for tat. But modern research shows precisely that this “surely” does not hold: in one-shot anonymous interactions treating others well is not repaid (Chapter 1); repayment appears only when the probability of meeting again is high enough (), reputations are known well enough (), or some other mechanism is at work. And even when there is repayment, it is conditional: tit for tat repays only those who repay, and the discriminators of Chapter 4 help only “good” individuals. Mozi’s “impartial care” is undiscriminating love, which in the models is closest to “always cooperate”, and always cooperate is precisely the strategy most easily exploited by defectors (Chapter 3). So if Mozi’s assertion about repayment is taken as an empirical proposition, modern research finds it true only under particular conditions; if it is taken as persuasion, what it seeks to persuade people of is that they should give first, which is a different matter.
Repay resentment with uprightness, repay kindness with kindness
Source text
或曰:「以德报怨,何如?」子曰:「何以报德?以直报怨,以德报德。」(杨伯峻 译注 1980)
(Analects, “Xian Wen”)
为无为,事无事,味无味。大小多少,报怨以德。(陈鼓应 注译 2003)
(Laozi, Chapter 63)
In the author’s translation: Someone said: “What about repaying resentment with kindness?” The Master said: “Then with what would you repay kindness? Repay resentment with uprightness, and repay kindness with kindness.” The Laozi says: Act without acting, do without doing, taste the tasteless. Whether great or small, many or few, repay resentment with kindness.
Original meaning and context. The phrase “repay resentment with kindness” also appears, as 报怨以德, in Chapter 63 of the Laozi, and the question in the Analects may have this view in mind, although the Analects does not say who asked. Confucius’s answer has two layers: if one repaid resentment with kindness, there would be nothing left with which to repay kindness, and failing to distinguish favour from grievance is itself unjust; resentment should be repaid with “uprightness”. The meaning of “uprightness” (直) has long been debated; Zhu Xi explained it as “in what one loves and hates, takes and rejects towards those one resents, being entirely impartial and without selfishness”: one’s dealings with someone one resents should come entirely from fairness, untainted by private grievance. This is not “an eye for an eye”, nor “ignoring the matter”, but treating the other as fairness requires.
Read beside modern research. Chapter 4 has already discussed this parallel. “Repay kindness with kindness” is close to “helping a good individual is good”; “repay resentment with uprightness” bears some resemblance to “justified refusal”: towards someone who has wronged you, you need not return kindness, but neither should you retaliate doubly out of private grievance. Generous tit for tat (Chapter 3) is formally related as well: it repays cooperation, and answers defection with retaliation with some probability and forgiveness with some probability, the limit on forgiveness being set by the payoffs. But the boundary is equally clear: Confucius is discussing the personal virtue of the gentleman, and “uprightness” is a disposition and a judgement whose standard is impartiality, not a table of strategies with stated probabilities; strategies in evolutionary models have no dispositions, and their “goodness” is measured by changes in their frequency in a population. To say that Confucius “proposed generous tit for tat” or “second-order reputation norms” would be inappropriate.
Three boundaries
Reading these ancient texts beside evolutionary game theory shows that their concerns overlap: should love and giving vary with the closeness of the relation? Is treating others well repaid? How should we treat those who wrong us? But three boundaries must be kept clear.
First, ought and is. Mencius, Mozi and Confucius discuss how people ought to act; evolutionary game theory discusses which strategies will be retained under given payoffs and structures of interaction. That a strategy is retained in evolution does not make it morally right; extortion strategies can prevail in small populations (Chapter 3), and that is no defence of extortion.
Second, ethical categories and mathematical quantities. “Love of kin”, “impartial care” and “uprightness” are ethical concepts of great richness, embedded in the historical contexts of the patrilineal order, ritual and the debates of the Hundred Schools; relatedness, the probability of continued interaction and reputation rules are precisely defined mathematical quantities. Between them only structural comparisons can be made; they cannot be translated into one another.
Third, the ancients did not “already discover” evolutionary game theory. Ancient thinkers noticed repayment and gradation in human relations, observations drawn from long experience of social life; writing payoffs as matrices, populations as frequencies, and calculating the fate of strategies with replicator equations and fixation probabilities are twentieth-century inventions. The value of reading the two side by side lies in seeing clearly what question each way of thinking answers, not in conferring modern titles on the ancients.
Chapter summary
“Love one’s kin, and so be humane to the people; be humane to the people, and so cherish creatures” in “Jin Xin I” of the Mencius is about the gradations of love; it resembles Hamilton’s rule in the form “give more to those who are closer”, but one is a ritual ethic and the other a genetic statistic. “Whoever loves others will surely be loved in return” in “Universal Love II” of the Mozi presents the repayment of impartial care as certain; modern research shows that repayment appears only under particular conditions and is itself conditional, and that undiscriminating love is, in the models, closest to “always cooperate”, the most easily exploited strategy. “Repay resentment with uprightness, repay kindness with kindness” in “Xian Wen” of the Analects distinguishes the response to favour from the response to grievance, and is structurally related to justified refusal and to generous tit for tat; but “uprightness” is a virtue measured by impartiality, not a table of strategies. The next chapter returns to practice.
References
Fan, Y. (2026). Evolutionary Game Theory, Direct and Indirect Reciprocity and the Emergence of Cooperation: From Hamilton's Rule and Axelrod's Tournaments to Reputation and Costly Punishment, Read Beside Classical Confucian and Mohist Ethics [演化博弈、直接与间接互惠及合作涌现的数学模型] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.23788551 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).