Si Wu Zi 《思無字》

A2 — Two Translations: Literal vs Performance

I

A familiar gesture repeats whenever a difficult text begins to matter. We praise it for “depth,” by which we usually mean that it resists quick paraphrase, and then we rush to paraphrase it anyway, as if the point of reading were to extract a portable conclusion and discard the object that produced it. When this fails—when the text refuses to be reduced without breaking—we reverse the compliment and call it “mystical,” “irrational,” or “merely poetic.” The two reactions are opposites only in tone. In structure they are identical. Both assume that a text is supposed to settle into a final statement, and both become impatient when it does not.

A1 was written against that impatience. It insisted on an unfashionable beginning: the text first. Si Wu Zi 《思無字》 was printed intact—four chapters, Chinese preserved, English definitions attached line by line—before any “interpretation” was allowed to take over the room. This was not a gesture of reverence. It was an elementary methodological requirement. If one claims to be reading a scripture, one does not hide it behind commentary and then congratulate oneself on the commentary.

But A1’s method has an immediate consequence, and it cannot be avoided. Once the text is in front of us, we are forced to ask what we are doing when we “stabilize” it into English. What is translation, in this project? Is it a neutral service—a kind of courier’s job, moving “the meaning” across a border? Or is it already a decisive act inside the system the scripture itself describes?

A close analysis shows that the neutral-service picture is misleading. Translation is not an external operation performed on Si Wu Zi. It is one of the text’s own operators. To translate is to work (作) in a new medium, to produce forming (形) that did not exist before, and then to leave a stabilizing trace—an authored version, a citeable line, a repeatable phrasing—that becomes 著 (authoring/stabilizing) in exactly the sense Chapter 2 makes central. A translation, once published, becomes an object people can return to, teach, quote, worship, or mistake for finality. That is not a peripheral problem. It is the book’s own ethical problem, named with precision in Chapter 3 as the passage from 假 (scaffold) to 妄 (premature closure).

This is why A2 exists. It begins from A1’s ground and asks the next unavoidable question: what happens to Si Wu Zi when you change its voice? What does a text that is written to prevent final landing look like when it is rendered into a language whose habits encourage landing? And what techniques are available to preserve—without romanticizing—what the scripture is doing?

The answer proposed here is deliberately modest and deliberately strict. We do not “solve” translation. We perform it in a controlled way. We translate key lines twice, in two distinct voices, and we force ourselves to see what each voice stabilizes, what each voice reopens, and what each voice risks freezing.

The first voice is Literal. “Literal” here does not mean naïve word-for-word substitution. It means rail-preserving translation. It attempts to keep visible the grammar that builds the system: connective engines like 而, 又; routing markers like 由 (“from/by/through”) and 至 (“to”); restart syntax like 「而著，而行，而作」; and the peculiar negation machinery of Chapter 1, especially 非非 (“not-not”), which functions less as skepticism than as a reopening device. Literal translation tends to sound strange in English, and that strangeness is a feature, not a defect. It is the cost of preserving rails that English would otherwise smooth into invisibility.

The second voice is Performance. “Performance” here does not mean decorative paraphrase. It means motion-preserving translation. It tries to make English behave like the scripture behaves. When Chapter 1 refuses final settlement, performance English must refuse settlement. When Chapter 2 loops, performance English must loop. When Chapter 2 phase-shifts at noon and midnight, performance English must feel like saturation and resurgence. When Chapter 4 functions as a score—taste, tone, pun, echo—performance English must be capable of carrying that sensory logic without converting it into mere “meaning.” Its highest virtue is not lyrical flourish. Its highest virtue is generating operational 信: a felt credibility that the system’s instability is real and livable. But that virtue carries a requirement: performance must force a return to the literal voice for structural verification, or else it becomes mere mood and therefore another form of stasis.

This pairing is not a stylistic indulgence. It is demanded by the scripture itself, because Si Wu Zi is built out of dualities that do not sit opposite each other like neat binaries. They twist. They behave like a Möbius strip: what looks like two sides becomes one continuous surface with a half-turn. If translation treats these dualities as simple oppositions, it will flatten the text into doctrine. If translation refuses to name them at all, it will dissolve the text into fog. The two-voice method is an attempt—imperfect, but workable—to keep the twist visible.

Before demonstrating the method chapter by chapter, one clarification is necessary, because it prevents a common misreading that would otherwise contaminate everything that follows. The key terms will be held steady. 相 is concept: the conceptual clamp that names and fixes. 象 is form in the noun sense: manifested configuration, the appearing form. 形 is forming: the verb-event of giving form, the act by which forms become. 行 is acting; 作 is working (production, the working-through of making); 著 is authoring/stabilizing as a verb (inscription, attribution, trace-making). And when we speak of “truth” or “meaning” here, we are not importing a theory from outside. We are following Chapter 3’s own insistence that coherence depends on 信—trust, credibility, transmissible reliability—rather than on any fantasy of final correspondence.

II

If the method were merely a matter of translating words, we could stop here. But Si Wu Zi does not permit that reduction, because it does not treat its key statements as “definitions” in the ordinary sense. It builds worlds through grammar. The most important example is the pattern 「X者 Y之 Z」, which English readers instinctively treat as a dictionary formula: “X is the Z of Y.” In Si Wu Zi, this is not a passive description. It is an installation of roles. It is a world-builder.

X is not a substance; it is a position. Y is not a possession; it is a generating field, a “world” in which X is being situated. And Z is not a simple predicate; it is often a dual or gradient construct that carries the Möbius twist. In other words, the sentence does not tell you what X “really is.” It declares how X functions as a role of Y, and it does so in a way that refuses final closure.

This matters most in Chapter 3’s triad:

「書者字之道法，言者文之相象，意者心之思念。」

A literal rendering, kept strict, would read:

“Writing (書) is the Way/Law (道法) of characters (字). Speech (言) is text’s concept-and-form (相象). Intent (意) is the heart’s thought-and-remembrance (思念).”

It is tempting to treat this as a tidy metaphysical chart. It is nothing of the sort. The hinge terms—道法, 相象, 思念—are not labels for objects. They are torsions: constraint and continuity; concept and form; projection and reconstruction. They are designed to prevent the reader from concluding, “Ah, writing is just this, speech is just that, intent is just the other.” The system is installed as a set of interlocked roles.

Now add the further insight you insisted on: the triad can be read as three “worlds,” but not stacked. They inter-contain cyclically, like a paper–rock–scissors topology. The “Platonic world” of shareable structure (speech as 文之相象) is a subset of the mental world, because ideal forms are generated and held by mind; the physical world of traces (writing as 字之道法) is a subset of the Platonic world, because physicality arrives already structured as form and pattern; and the mental world is a subset of the physical world, because mind is embodied and realized in substrate. No layer can exhaust the others, because each depends on what it tries to ground. Read this way, the famous insufficiency cascade is not a lament about expressive failure. It is a topological statement:

「書不盡言，言不盡意，意亦然不盡字。」

Writing cannot exhaust speech; speech cannot exhaust intent; intent likewise cannot exhaust characters. Not because we are tragically limited, but because the layers loop, inter-contain, and twist. Every attempt to make one layer sovereign is structurally defeated. The system keeps moving.

If this is correct—and the point of the essays that follow is to test it, not to declare it—then translation cannot be a neutral conduit. It is necessarily a new stabilization event inside a loop that treats stabilization as both necessary and dangerous. Translation is therefore ethically charged in the book’s own terms. A translation is a 假—a scaffold that makes return possible. It becomes 妄 the moment it pretends to be the final form.

With that in mind, we can now perform the demonstrations.

III

Chapter 1, 「言無言」, is the gate. It installs the anti-freeze constraint. If translation fails here, the entire project collapses into one of two familiar mistakes: either mystical resignation (“it cannot be said”), or philosophical slogan (“language fails”). Both are forms of stasis. Both betray the chapter’s actual operation.

The core passage reads:

「言無言，言，相之表述，又言，語裡之象，於象離相；見眾象非相，即見如來，無相而象無言，非非相，非非象，非非言。」

A literal translation, kept rail-preserving, would proceed phrase by phrase:

「言無言」 — “Speech is not speech.”
「言，相之表述」 — “Speech: the articulation of concept (相).”
「又言，語裡之象」 — “And again speech: the form (象) inside utterance.”
「於象離相」 — “Within form, detach concept.”
「見眾象非相」 — “See the many forms are not concept.”
「即見如來」 — “Then one sees the Thus-Come (Tathāgata).”
「無相而象無言」 — “Without concept, form has no (final) speech.”
「非非相，非非象，非非言」 — “Not-not concept; not-not form; not-not speech.”

This version has the virtue of being accountable to the Chinese. It preserves the doubled “speech” that is essential to the chapter’s strategy: language has a conceptual function (相之表述) and a form-bearing function (語裡之象). It preserves the crucial instruction 「於象離相」, which is not an escape from form but a de-gripping within form. It preserves the anti-absolutism of 眾 (“many”), which prepares the later claim 「象無一象」 (“form has no single form”). And it preserves 非非 as a reopening mechanism rather than reducing it to a vague “neither/nor.”

But literal translation also carries a danger, precisely because it stabilizes. “Speech is not speech” can easily be read as a proposition to admire, or as a mystical claim to fetishize. “Not-not concept” can be treated as a clever trick. The gate becomes an object. The reader lands.

A performance translation exists to prevent that landing. It must make English behave like the gate behaves. A performance rendering might read:

Speech that refuses to become a home.
Speech that pins concept—and speech that carries form inside what is said.
Stay in the form; loosen the concept’s grip.
See forms in their many configurations: no concept is identical with them.
Then seeing arrives without seizure.
Without the clamp, form will not be finished by what you can say.
And do not make negation your refuge: not-not concept, not-not form, not-not speech—leave every closure reopenable.

Here “Thus-Come” is handled functionally, not doctrinally: it names a mode of seeing that does not collapse form into concept. The point is not to import a religion; it is to mark a perceptual stance. Likewise, “not-not” is not treated as a logical flourish but as an operational safeguard: you do not cure fixation by declaring everything unreal, because that too becomes an idol.

What has changed? Not meaning in the shallow sense, but what becomes stable. Literal voice stabilizes the internal anatomy. Performance voice stabilizes the discipline. Both are needed, because Chapter 1 is fighting two symmetrical temptations: the temptation to freeze meaning into concept, and the temptation to “solve” freezing by retreating into negation. It is not enough to state that this is happening. The translation must enact it.

IV

Chapter 2, 「行有形」, is the engine. It makes explicit what Chapter 1 implied: that the system is not a set of claims but a set of operators. It also makes explicit something most translation discourse prefers to deny: that medium-shift is not “delivery of content” but continued making.

Consider the chapter’s opening moves:

「行有形，行作而形著，而著，而行，而作；有著，有行，有作。」

Literal translation, kept strict, must preserve the loop grammar:

「行有形」 — “Acting has forming (形).”
「行作而形著，而著，而行，而作」 — “Acting works; forming is authored/inscribed (著); and once authored—acting again, working again.”
「有著，有行，有作」 — “There is authoring; there is acting; there is working.”

If English smooths this into a tidy summary—“action creates form which is then recorded”—the loop is lost. The “and once authored—acting again” is not decorative repetition. It is restart logic. 著 is not closure; it is hinge.

The chapter then deepens the feedback:

「作中之著則行旺，著中之作則形盈。」

Literal translation:

“Authoring within working makes acting flourish; working within authoring makes forming full.”

This is not metaphor. It is a statement about nested operations: stabilizing trace inside production amplifies action; production inside trace deepens formation. A translation that renders 著 merely as “attachment” or “ego” destroys the mechanism. Here 著 is structural, not moral.

Now the chapter’s phase-shifts:

「著形之午，形由作降；作行之子，行由著升。」

Literal translation:

“At the noon of authoring-forming, forming descends from working; at the midnight of work-acting, acting rises from authoring.”

Noon and midnight here are not poetic scenery. They are control logic: saturation yields decline; decline becomes the condition of resurgence. And note the repeated 由 (“from/by/through”), which is routing, not ornament: form descends from work; action rises from trace.

Then latency:

「作先行，後浮出作；著先形，後沉入著。」

Literal translation:

“Working goes first with acting, later the working floats up; authoring goes first with forming, later it sinks into authoring.”

This is a maker’s observation elevated into system law: causes hide before effects surface; traces become infrastructure.

Now the cross-media operators that make A2’s thesis impossible to ignore:

「以行煮以話形，著畫形以作聞，又以著行以聲作。」

Literal translation:

“Use acting to cook forming into speech; use authoring to draw forming so as to make hearing; again, use authored acting to work through sound.”

If one still wants to claim that translation is a neutral service, one must explain why the scripture itself treats speech and hearing as produced through cooking and drawing—through operations, not through transfer. A translation is precisely such an operation. To translate is to cook forming into another speech; to draw forming into another hearing; to work through another sound system. Translation is not outside the engine. It is the engine.

Finally, the closing line you flagged:

「右左依著遊。」

Literal translation:

“Right and left, it roams relying on authoring/stabilization.”

But the literal meaning is only part of the line’s function, because 右左 carries a punning echo of 由作—roaming “by/through making.” The line says: roaming depends on stabilization (依著), and roaming routes through production (由作). It is almost a compressed description of the translator’s proper posture: move, but rely on trace; improvise, but route through work.

A performance translation must make the loop felt:

Wherever there is acting, forming is already happening.
Acting enters work; forming leaves trace; the trace does not end the process—it forces return.
At noon, when the authored form is most visible, formation sags under its own frame; at midnight, from the same stable trace, action rises again.
Work exists before it appears; authorship begins loudly and then disappears into infrastructure.
And making does not stay in one medium: forming can be cooked into speech, drawn into hearing, worked through sound.
So the process roams right and left—by making, through making—on the rail of its traces: moving because it is stabilized, alive because it will not freeze.

Here the point is not to decorate the text. The point is to preserve what the chapter is doing to the reader: forcing the recognition that stabilization is not the enemy but the hinge, and that the hinge becomes a prison only when it is mistaken for an end.

V

If Chapter 2 makes translation-as-operator plausible, Chapter 3 makes it ethically unavoidable. Chapter 3 tells us why stabilization is required and why it produces delusion.

Start with the triad again:

「書者字之道法，言者文之相象，意者心之思念。」

Literal translation, with the “three worlds” cue kept explicit:

Writing is the Way/Law (道法) of characters (字). Speech is text’s concept-and-form (相象). Intent is the heart’s thought-and-remembrance (思念).

Now interpret this in the topology you insisted on—not as a hierarchy but as a cycle of mutual containment. The “Platonic” (speech as shareable structure: 文之相象) is generated within the mental (意: heart’s thought and remembrance). The physical (writing as trace and constraint: 字之道法) is structured by the platonic (forms and patterns). And the mental is realized within the physical (embodied substrate). Each world both contains and is contained, depending on how you traverse the strip. That is why no “definition” can close.

Now consider the core proof-chain:

「文，行象于字，依法以書畫，象無一象，妄象假念。」

Literal translation:

“Composition/text (文) moves form (象) through characters (字); by law it writes and paints; form has no single form; delusive form arises from provisional remembrance (假念).”

If 相 is concept and 象 is form, then 「象無一象」 becomes decisive: form is intrinsically non-singular. It is not merely that there are “many images.” It is that form cannot be sealed as one. The moment a “one” appears, it is an imposition. And the imposition is not random. It is built from memory-work: 假念. The scaffold becomes a claim of finality. The result is 妄象: delusive form, not because form is bad, but because closure is premature.

A performance rendering must preserve the inevitability:

Composition operates through tokens under constraint: it writes, it paints, it produces manifested forms—plural by principle.
Because form will not be one, concept keeps trying to force a one anyway.
Remembrance builds scaffolds that allow return, then mistakes the scaffold for the thing itself.
Delusion is not error in the ordinary sense. It is premature closure: forced singularity stamped onto what cannot be singular.

Now add the trust line:

「文，形相于字，知道以來信，相不足相，相不思過。」

Literal translation:

“Composition shapes concept (相) within characters; knowing Dao, trust/credibility (信) arrives; concept is insufficient to be concept; concept does not overstep.”

This is where translation ethics crystallizes. A translation is a shaping of concept in characters—only now the characters are English. It creates a stable phrasing that people can cite. That stability requires 信: not belief in dogma, but operational trust that the scaffold is workable. Yet concept is insufficient; it cannot close what it claims. If the translator (or reader) forgets this, concept oversteps—becomes sovereign—and the translation becomes a doctrine. That is the very stasis Chapter 1 was designed to prevent.

Finally, the cascade:

「書不盡言，言不盡意，意亦然不盡字。」

Literal translation:

“Writing cannot exhaust speech; speech cannot exhaust intent; intent likewise cannot exhaust characters.”

The three-world loop makes the meaning sharper. Each layer fails to exhaust the next because each depends on the others. A translation that pretends to be exhaustive is not merely inaccurate. It is structurally incoherent with the scripture it claims to render.

The chapter closes with the reader’s discipline:

「念一字，思無字。」

Literal:

“Hold one character in remembrance; think without characters.”

This is the proper stance toward translations themselves. Hold one version firmly enough to work with it—to return, to verify, to build. Then refuse to idolize it. Reopen beyond it. Make the scaffold a hinge, not a prison.

VI

If the preceding chapters build the mechanism, Chapter 4, 「心由幸」, shows what the mechanism feels like when lived. It is the point at which translation most commonly fails, because translators often assume that the poem is an “appendix,” a decorative flourish after the “real philosophy” is done. The poem itself refutes this assumption by behaving like a score: taste, tone, step-count, return, response, cooking, pun, echo. If one translates it as a set of paraphrasable propositions, one commits the sin the book anatomizes: one collapses form into concept.

The poem reads:

「沾苦淡閒言
甘問宮
幾步多
路回足
自行走
近圓空
似作達而意煮
辛油形即答心」

Literal translation must keep sensory roles intact:

“Touched by bitterness, idle words become thin.”
“Sweet (甘, taste) questioning meets Gong (宮, the gong tone; 五音 resonance in Five-Phase correspondences).”
“How many steps—so many.”
“The road answers back to the foot.” (路回足 is not merely “the road returns”; it is response: the road returns an answer to the foot.)
“Self-walking / walking by oneself.”
“Near the round emptiness.”
“It seems working reaches, yet intent is cooked.”
“Pungent-oil forming immediately answers the heart.” (辛油 is pungency and medium, and it puns toward 幸; grace is carried as taste and sound, not as concept.)

A performance translation must preserve the poem’s function as lived recursion:

Bitter touch thins the chatter.
Sweetness is tasted; Gong is sounded—questioning becomes tuning.
Steps multiply without arrival.
The road answers back to the foot: walking learns response, not conquest.
The loop is carried by the body—self-walking, not theory.
Near the round empty: orbit, do not possess.
What looks like attainment is cooking; intent is prepared in heat.
And then—pungent oil, grace by sound—forming answers the heart at once.

Here the poem is not “explaining” the system. It is enacting it. It thins idle speech (the cost of insufficiency), tunes taste and tone (meaning as calibration), counts steps without finality (anti-landing), receives response from the road (recursion as dialogue, not repetition), nears a round emptiness (non-closure as lived geometry), cooks intent (work as production of meaning), and ends with a punning grace (幸) carried through medium (油) and sensation (辛). The poem returns the reader to Chapter 2’s engine by another route: not diagram but body.

VII

At this point the lesson is not difficult to state, though it is difficult to practice. Literal voice reveals rails and protects accountability. Performance voice reveals motion and protects the work’s status as artwork. Each voice hides something the other reveals. Literal can become dead architecture, worshipped as final. Performance can become free atmosphere, drifting into uncheckable feeling. Either failure mode is already named in Si Wu Zi’s own vocabulary: the first is 著 mistaken for end; the second is 信 without verification, which becomes another kind of stasis.

The scripture itself tells us how to hold the tension. Chapter 3 treats scaffolding (假) as necessary; it becomes delusion (妄) only when the scaffold claims finality. Chapter 1 prevents the opposite mistake by refusing the idol of negation: 非非 reopens rather than denies. Chapter 2 makes stabilization the hinge of return rather than the tombstone of completion. Chapter 4 demonstrates that the engine, when lived, is not a doctrine but a rhythm: bitterness thinning speech, sweetness tuning inquiry, the road answering back to the foot.

This is why A2’s “two translations” are not an academic game. They are a reading discipline demanded by the object. Read in literal voice to identify the rails: the world-building grammar, the connectives, the routing markers, the restart commands, the Möbius hinges embedded in hinge-terms like 道法, 相象, 思念. Then read in performance voice to feel what the rails do: refusal of landing, looping, phase-shift, latency, sensory tuning. When performance begins to feel like fog, return to literal. When literal begins to feel like a dead diagram, return to performance. This oscillation is not optional if one wishes to avoid the two temptations the text is written to defeat: the temptation to freeze meaning into doctrine and the temptation to dissolve meaning into mystique.

A2 therefore ends, as it should, not with a conclusion but with a capability: dual-voice reading as practiced oscillation, anchored by the scripture’s own seal.

「念一字，思無字。」

Hold one character in remembrance; think without characters. Commit to a token so you can return—then refuse to idolize the token so the field can reopen. That instruction is not only a metaphysical slogan. It is translation ethics. It is the rule by which every future essay in this project must be judged: whether it leaves the text runnable without turning it into a statue.