International Institute of Jesus Christ
John 2
Water becomes wine at the wedding in Cana. Jesus goes up to Jerusalem and drives the traders out of the temple.
Why this rendering
John.2.4 Hypothesis
Syntactic observation
Τί ἐμοὶ καὶ σοί contains no verb. It is one nominative neuter interrogative (Τί) and two dative pronouns (ἐμοί, σοί) joined by a conjunction. γύναι is vocative, and the narrator calls Mary μήτηρ (mother) at 2:1, 2:3, 2:5 and 2:12 — the switch to γύναι happens only in Jesus' own mouth. ὥρα has already appeared at 1:39, but there it reads a clock — ὥρα ἦν ὡς δεκάτη, “it was about the tenth hour.” Here it carries the article and a possessive genitive, ἡ ὥρα μου, Jesus’ own hour. That use begins here. The Greek carries no morphological marker of the speaker's register or deference. SBLGNT marks a variant (⸀) at the opening καί.
Rendering options · 4
Basis Renders the idiom as a question of concern, and Korean adds an honorific ending.
Rejected The symmetry of the two datives collapses into a single notion of ‘concern’, and Korean introduces a register the Greek does not mark.
Basis Renders what the idiom is doing pragmatically rather than what it says.
Rejected It invents a verb the text does not have. The strangeness of the idiom is a property of the text, not noise to be removed.
Basis Treats γύναι as too harsh for a mother and substitutes the kinship term.
Rejected The narrator uses μήτηρ and Jesus uses γύναι — that switch is a fact of the text. Restoring ‘Mother’ erases it. The Chinese Union Version appends a note, 「原文是婦人」, admitting the substitution.
Basis Keeps the two datives in symmetry and leaves γύναι as a vocative. ὥρα is rendered ‘hour’ so the recurring motif is visible. Korean must choose a register; we follow convention and record the forced choice under lost_in_translation.
Lost in translation
The Greek juxtaposes two datives with no verb. Korean and English must supply one, and supplying it fixes the relation.
Greek marks no deference level. Korean must encode one in the verb ending, so any choice adds information the Greek does not carry. No translation escapes this.
The narrator says μήτηρ, Jesus says γύναι. Rendering Jesus' word as ‘mother’ merges the two.
Why this rendering
John.2.11 Observation
Syntactic observation
ἀρχήν is a noun, not an ordinal. It is accusative singular, and it is the same word (ἀρχή) that opens the Gospel at 1:1, ἐν ἀρχῇ. σημείων is genitive plural, so the phrase is ‘a beginning of the signs’ — the start of a series, not a count of one. SBLGNT marks a variant (⸀) at ἐποίησεν.
Rendering options · 2
Basis Reads ἀρχήν as an ordinal for a smoother sentence.
Rejected Turning ἀρχή into ‘first’ hides that it is the same word as ἐν ἀρχῇ in 1:1. The reader cannot see that the Gospel opened with it.
Basis Keeps ἀρχή as a noun so that its identity with 1:1 remains visible, per the corpus-consistency rule in GLOSSARY.
Lost in translation
ἐν ἀρχῇ at 1:1 and ἀρχήν at 2:11 are the same word. An ordinal makes them two different words and severs the link.
Why this rendering
John.2.16 Observation
Syntactic observation
οἶκον occurs twice in one sentence. Both are accusative singular of οἶκος — the same word. The first is ‘the οἶκος of my Father’, the second ‘an οἶκος of ἐμπόριον’. The psalm quoted in v.17 uses the word again, τοῦ οἴκου σου — three occurrences across two verses. μὴ ποιεῖτε is a present imperative with the negative.
Rendering options · 2
Basis Turns ἐμπορίου into a place-noun for readability.
Rejected The second οἶκος disappears and the ‘house … house’ repetition breaks. The saying sets one word against itself; replacing one side with ‘market’ turns a contrast into a substitution.
Basis Keeps ‘house’ for both, so the repetition carries through to ‘your house’ in the citation at v.17.
Lost in translation
Twice in v.16 and once in the citation at v.17 — the same word three times. Rendering one as ‘market’ breaks the chain and hides that the citation picks up the very word Jesus used.
Why this rendering
John.2.19 Hypothesis
Syntactic observation
The chapter uses two different words for the temple. Verses 14 and 15 have ἱερόν (the neuter of ἱερός in the morphology); verses 19, 20 and 21 have ναός, a masculine noun. They are distinct words with distinct roots, and the text switches between them inside a single chapter. The traders are driven from the ἱερόν; what is to be destroyed is the ναός. Verse 21 places that ναός in apposition with τοῦ σώματος αὐτοῦ — his body — and so names what was meant. What the morphological data establishes is that the two words differ; that ἱερόν denotes the precinct and ναός the dwelling proper is lexical knowledge. We have not checked a lexicon, so this unit is published as a hypothesis.
Rendering options · 2
Basis Collapses both words into one superordinate term for readability.
Rejected ἱερόν at v.14 and ναός at v.19 become the same word, so the building Jesus cleansed and the thing he tells them to destroy read as one object. Verse 21’s “his body” then arrives as an abrupt figure rather than the resolution of a shift the text has already made.
Basis ἱερόν is the whole precinct; ναός is the dwelling proper. Distinguishing them keeps in the sentence the fact that v.19 points to something other than the courts just cleansed. σκηνόω at 1:14 — to pitch a tent — already pointed to the same thing: the place where God dwells. NET, Elberfelder and the Japanese Shinkyōdōyaku each preserve the distinction in their own way.
Lost in translation
The original uses two words; most versions use one. 개역한글, ESV and NASB have 성전 / temple; the Vulgate templo / templum; Luther Tempel / Tempel; the Synodal храме / храм. Only three keep it: NET renders vv.14–15 “temple courts”, the Japanese Shinkyōdōyaku uses 神殿の境内, and Elberfelder glosses the two in parentheses as “die Gebäude” and “das Heiligtum”.
Why this rendering
John.2.24 Observation
Syntactic observation
ἐπίστευσαν in v.23 and ἐπίστευεν in v.24 are the same verb, πιστεύω. Many πιστεύω-ed in his name; Jesus did not πιστεύω himself to them. The tenses differ — aorist in v.23, imperfect in v.24. Verse 24 carries two SBLGNT variant marks — the particle ⸀δέ and the pronoun ⸀αὑτόν. The second is the reflexive (ἑαυτοῦ) and bears on the sense. A single breathing mark separates the reflexive αὑτόν from the plain pronoun αὐτόν.
Rendering options · 2
Basis Renders the second πιστεύω as “entrust” to make the sense plain.
Rejected “Believed” in v.23 and “entrust” in v.24 become different words, and the parallel the author built by repeating one verb disappears. They believed in him; he would not believe himself over to them — that turn is the shape of the sentence.
Basis Korean can share the root by saying “believe and hand over”. English has no single verb, and keeping πιστεύω = believe as GLOSSARY requires makes the echo unrecoverable; that asymmetry is recorded below. The imperfect is read as a sustained stance rather than a single act.
Lost in translation
Verses 23 and 24 set the same verb against itself. English splits it into believe and entrust and the parallel breaks; Korean likewise splits 믿었으나 / 의탁지. Latin keeps crediderunt / credebat, Russian keeps the вер- root in уверовали / вверял, and Japanese shares 信 in 信じた / 信用.
The aorist in v.23 points to what happened; the imperfect in v.24 to a stance that continued. Rendering both as simple past erases the contrast. YLT marks the imperfect with “was not trusting”.