The first line of the Tao Te Ching is punctuated two ways, and the choice decides what the whole book is about.
One reading:
The nameless is the beginning of heaven and earth; the named is the mother of ten thousand things.
Always without desire, observe its mystery; always with desire, observe its boundaries.
The other:
Non-being names the beginning of heaven and earth; being names the mother of ten thousand things.
Always hold non-being as your frame, and you see its wonder; always hold being, and you see its outline.
Every translation you have read takes the first. I take the second, and so did a line of Song-dynasty commentators beginning with Wang Anshi.
For years the standard objection ran like this: the Mawangdui manuscripts (c. 168 BCE) read 恒无欲也,以观其妙 — with a sentence-final 也 after 无欲. That 也, the argument goes, closes the phrase "without desire," so the two characters 无欲 must be read together. Case closed.
It is not closed. And the reason is a small piece of Classical Chinese grammar that gets skipped.
也 closes a clause, not a word
In Classical Chinese, sentence-final 也 marks a full judgment. It does not attach to the two or three characters immediately before it; it closes the predicate.
So 恒无,欲以观其妙也 parses perfectly well as:
- 恒无 — a frame, held constantly ("always non-being")
- 欲以观其妙 — the main clause ("one intends thereby to observe its wonder")
- 也 — closing the whole thing
The manuscripts have no punctuation. They were copied as continuous strings of characters, and every punctuation mark you have ever seen in a Laozi text was added by a later editor, working from a later reading. Continuous writing is not the same as bound syntax. A character appearing next to another character proves adjacency, not constituency.
The 也 is real evidence. It is evidence that the reading "always without desire" is natural — it flows, it is what a reader expects, and it has held for nearly two thousand years. It is not evidence that the other reading is impossible.
That distinction — natural versus impossible — is where most of this debate has quietly gone wrong.
Why the punctuation matters more than it looks
If 无欲 is a fixed phrase, then 欲 means desire, craving, the pull outward. Non-being becomes a state of desirelessness, being becomes the world of desire and differentiation, and the path is subtraction: quiet the craving, still the mind, return to the clean place.
That reading has real depth. It also has a cost: it turns chapter 11 — "being provides the benefit, non-being provides the use" — into a hierarchy. Non-being becomes the pure level, being the fallen one. And once that hierarchy is installed at chapter 1, it runs through all eighty-one chapters.
The other reading makes 欲 a vector rather than an appetite: the mind's capacity to aim, to intend, to look. Not good, not bad — just directional. You can aim your attention at the frame "non-being" or at the frame "being," and each shows you something the other cannot.
Under that reading, 玄 — usually translated "mystery" — is not a higher realm above the pair. It is the state of holding both frames at once without settling on either. Two threads, twisted together. That is literally what the character shows: 玄 is a pair of silk threads interwoven.
And 妙 / 徼 stop being ranked. 妙 is what has not yet taken shape; 徼 is the outline that lights up once it does. Same substance, two aspects, no hierarchy.
The honest position
I do not claim this is the one punctuation Laozi intended. Nobody can claim that. The manuscripts have no marks; the marks are ours.
What I claim is narrower: the 也 does not settle the question, the two readings produce two coherent but incompatible philosophies, and a book ought to pick one and stay inside it rather than drifting between the two chapter by chapter.
The reading I follow has its own lineage — Wang Anshi, Lü Huiqing, Su Zhe, and in the twentieth century Gao Heng, who argued it on textual grounds. It is supported by a line in the Zhuangzi: 建之以常无有 — "established on constant non-being and being," listed as a pair, not as a sequence.
The other lineage is equally serious. Zhu Xi rejected the split reading outright, on the grounds that it simply does not read smoothly. He was right about the smoothness. He was not right that smoothness decides.
What this is, in the end
A grammar note that changes the first line, and therefore the frame of the whole book. Not a correction of the tradition — a choice made inside it, declared openly, and then held consistently for eighty-one chapters.
The full Chinese reading — 81 chapters, 3 supplementary essays, 11 appendices, roughly 886,000 characters — is at github.com/q1z2q3-debug/runze-daodejing. The version discussed here (v2.3.0) adds an appendix devoted to exactly this argument.
Author: 润泽博士 (Runze) · #润泽博士
Top comments (0)