Conclusion

This article asked whether not adding a control, or withdrawing a promise, is also a restraint, and who recognises it. Its contribution is to bring the Dao De Jing, read at length in Legge's English, into the debate over rule design, and to add subtraction as restraint to that debate.

Against a design side that asks for rules well made, and a residue and drift side that watches rules accumulate and change under unchanged wording, the text locates the harm of rules in their number and in the activity of those who make them. It describes a way of governing that subtracts: diminish and again diminish, promise little, act while things are small, know when to stop. It also refuses to dispense with correction, and in chapter 79 keeps the record of an engagement without pressing it. The synthetic scene shows a subtraction, a pause and a removal, and the platform's readings show what a subtraction needs if it is not to overreach or to harden, a boundary, a deadline and a reason handed on, and who carries the cost of these. On recognition the two part: the text praises a restraint no one sees, the readings ask that those who wait under it see its scope. The text shows what invisible restraint gains; the readings show what it risks, a removal no one can explain and labour no one counts.

The limits are those of Section 7.5: one translation, read without the Chinese; one synthetic scene, much of whose reasoning follows a common template; moderate agreement in the ledger. The article does not judge whether the wording should have been removed.

Further research could go three ways. A textual study could read the chapters used here in the Chinese, and ask how much of the restraint's shape, especially its qualification as action "(of purpose)" or "(with an ulterior purpose)", is in the text and how much in Legge's brackets. A study of the parent study's other scenes could ask whether the platform answers other subtractions, such as a withdrawn promise or a stopped flow, with the same bounded additions, which would show whether the response is a property of the template. And an empirical study, which would need working records rather than synthetic scenes, could ask whether removals recorded with their reason and condition for return are reversed more readily, or more appropriately, than removals left without a trace. Neither the text nor the scene can answer that.

Appendix: Ledger Figures Used

All figures are from the parent study's numbers file, from the coding made without the studio's frame, in its terms.