Bridging Divides: How 'Shu' from The Analects Can Harmonise Modern Relationships — Epoche C1
The one word In Analects 15.24 — numbered 15.23 in some editions — the disciple Zigong asks Confucius whether there is a single word that can be practised for a whole life, and the answer is shu 恕, glossed immediately as: what you do not want for yourself, do not do to others (己所不欲，勿施於人). The same maxim is given to a different disciple at 12.2. This essay is about whether that word does the work commonly asked of it — defusing an argument in a comment thread, keeping a family group chat civil — and the answer is that it does, but not by the mechanism usually attributed to it, and not with the ease usually assumed. Three things have to be established before any of that can be argued. The first is that shu is half of a pair and does not stand alone. The second is that it is not empathy, and calling it empathy imports a psychological operation the text does not require and which does not work. The third is that Confucius himself told his most able student that the student had not attained it — which is difficult to square with the recommendation that a reader adopt it consciously and immediately. Shu is half of something At Analects 4.15 the Master remarks to the disciple Zeng that his way is threaded on a single strand. After he leaves, the other students ask Zeng what was meant, and Zeng answers that the Master's way is zhong shu 忠恕 and nothing else. The pairing is not decorative. Zhong 忠 is usually translated loyalty, but the sense in this compound is doing one's utmost — a full and unstinting discharge of what one's own commitments require. Shu is the outward-facing member: extending from one's own case to another's. Split the pair and each half fails in a characteristic way, which is why the omission in the earlier version of this essay was not a matter of completeness. Shu without zhong is accommodation without content: a disposition to refrain from imposing, held by someone who has no position of their own to refrain from imposing. That is not the Confucian virtue; it is conflict-avoidance, and it is exactly what the essay's earlier advice — value the family bond above the disagreement — describes if nothing else is added. Zhong without shu is the familiar figure of the person of principle who is impossible to live with. The recommendation the text actually makes is to hold a demanding standard for oneself and to decline to impose it on anyone else, and both clauses are load-bearing. Shu is not empathy The earlier version described shu as projecting oneself into another's shoes and imagining their feelings. That is a modern psychological operation, and the text describes something else — a procedure of moral analogy. The textual evidence is at Analects 6.30, where the positive counterpart of the maxim appears: the person of ren — the cardinal Confucian virtue, variously rendered benevolence, humaneness or consummate conduct — wishing to establish himself, establishes others, and wishing to advance, advances others. Confucius then names the method: being able to take what is near at hand as the analogy (能近取譬). What is near at hand is one's own case. The procedure is to reason from the treatment one would find acceptable oneself to the treatment appropriate for another, and it is a piece of practical inference, not an act of imagination about someone else's inner states. David Nivison (1996) set out the logic of these arguments in Chinese moral philosophy and showed how much turns on the direction of the extension: the reasoner uses himself as the standard case and projects outward, which is a very different operation from attempting to reconstruct the other person's mind. Herbert Fingarette (1972) pressed the point further, arguing that the Analects is not a work about inner psychological states at all — that its subject is conduct within ritual and role, and that reading it as an account of feelings and intentions is an anachronism imported by readers formed on a different tradition. One need not accept Fingarette's whole position to accept the negative half of it, which is well supported: nothing in the passages defining shu asks the practitioner to feel anything. This is not a pedantic distinction, because the two operations have different failure modes and, as it happens, very different track records. What happens when people take another's perspective The essay's earlier prescription was to pause and ask oneself: if I were in their position, how would I read this remark, and what might they be worried about. This is perspective taking in the ordinary sense, and it has been measured, extensively, with results that do not support the prescription. Nicholas Epley, Boaz Keysar, Leaf Van Boven and Thomas Gilovich (2004) proposed and tested a mechanism for how people do it. Their account is that perspective taking is an anchoring and adjustment process: one begins from one's own perspective, which is automatically available, and then adjusts away from it towards what the other person might think — and, as with anchoring generally, the adjustment stops early, typically at the near edge of the range of plausible values rather than at its centre. Their evidence includes the pattern one would predict if this were the mechanism: adjustment is more complete when people are given time and incentives, and less complete when they are hurried or under cognitive load. The implication for the recommendation is direct. Instructing an angry person to imagine the other's view, at the moment of anger, is instructing them to perform an effortful adjustment under precisely the conditions that truncate it. What they will produce is their own view, slightly moved. The corresponding error in the other direction was documented much earlier. Lee Ross, David Greene and Pamela House (1977) asked students whether they would walk around campus wearing a sandwich board carrying a slogan, and then to estimate what proportion of their peers would agree. Those who agreed estimated that most others would agree; those