Executive Summary — Philemon (English → Cantonese)

Why it matters

Philemon is a single 25-verse letter, but it compresses an unusually high density of Critical and High doctrinal risk into a very small space — and it does so for a Hong Kong Cantonese audience with almost no existing colloquial-Cantonese translation precedent to lean on (unlike Romans). The letter’s central move — declaring a slave a “beloved brother” to his own master “in the Lord” — sits directly on top of three live Hong Kong cultural fault lines: the historical 妹仔 (mui-tsai) bonded-servant system, the contemporary 外傭 (foreign domestic helper) labor discourse, and the sworn-brotherhood (結拜) tradition of Cantonese popular culture. Getting this book right requires the same rigor as Romans, applied to a shorter but more concentrated text.

Key findings

• 21 doctrines identified across the full letter (1:1–25, every verse reviewed, none silently skipped): 3 Critical, 9 High, 6 Medium, 3 Low — an unusually top-heavy risk profile for a 25-verse book.
• 59 glossary terms now governed by the extended translation_memory.json (14 inherited exactly from the Romans baseline + 45 new Philemon-specific terms), plus 11 established proper-noun transliterations.
• Term-level risk distribution across all 59 governed terms: 6 Critical (all inherited: 神, 主, 耶穌, 基督, 恩典, 聖靈), 10 High (2 inherited: 信心, 聖徒; 8 new: appeal/懇求, voluntary/自願, slave/奴僕, brother/弟兄, charge-to-account/算落…嘅數上, situational obedience/順服, graciously restored/蒙恩歸還, human spirit/心靈), 22 Medium, 21 Low.
• The core passage (Philemon 1:8-21) is the only span where all three Critical doctrines (Grace, Lordship of Christ, Forgiveness and Reconciliation) overlap with nearly every High-risk doctrine simultaneously — it must be reviewed as one rhetorical unit, not fragmented segment-by-segment.
• Philemon introduces zero new Critical-tier terms but a dense cluster of 8 new High-tier terms, several of which (奴僕, 弟兄, 算落…嘅數上) sit inside “crowded semantic neighborhoods” where Cantonese offers too many tempting native alternatives rather than too few — the opposite failure mode from typical translation gap analysis.
• Two Greek wordplays on the name Onesimus (1:11, 1:20) are structurally untranslatable through the established Cantonese transliteration (阿尼西母) and require mandatory translator notes rather than invented phonetic solutions.

Risks

• Highest risk: δοῦλος + ἀδελφός at 1:16 (“no longer 奴僕 but… 弟兄”) — softening either term toward its nearest cultural neighbor (僕人/外傭-flavored “worker,” or 結拜-flavored “chosen peer alliance”) would silently dissolve the verse’s entire rhetorical shock, the letter’s doctrinal center.
• Grace-family inclusio (1:3, 1:22, 1:25) — χαρίζομαι at 1:22 must visibly share the 恩 root with 恩典, or Paul’s hoped-for release risks reading as prayers “earning” his freedom — precisely the Wong Tai Sin-style transactional exchange the baseline Critical grace doctrine forbids.
• ἐλλογάω at 1:18 (“charge it to my account”) — Hong Kong’s financial-hub register makes this metaphor feel too natural; the risk is not obscurity but flattening a substitutionary-imputation illustration into ordinary secular bookkeeping.
• ἑκούσιον/voluntary (1:14) and ὑπακοή/obedience (1:21) — both collide with Confucian filial-duty (孝順) and folk-religious vow-discharge (還神還願) obedience models; either collapse would misread grace-motivated response as owed debt.
• πνεῦμα at 1:25 — referent-confusion risk in both directions: 聖靈 imports the wrong (divine) referent, bare 靈 imports folk-mediumship associations.
• 家庭教會 at 1:2 — the single most easily-missed anachronism risk: a natural-sounding modern compound that imports an unintended contemporary Mainland political/legal connotation onto an ordinary first-century meeting arrangement.

Opportunities

• Because Philemon has thin existing colloquial-Cantonese precedent, this curriculum’s term decisions (奴僕, 懇求, 蒙恩歸還, consistent 心/心腸) are genuine original contributions, not mere confirmations — an opportunity to set a corrected standard rather than inherit existing translation drift (e.g., existing Chinese Bible traditions inconsistently render σπλάγχνα across 1:7/1:12/1:20; this package fixes that).
• The letter’s warm, personal register (σπλάγχνα, ἀγάπη, χαρά vocabulary) offers a natural bridge to teach Romans 8/12-style relational warmth within a short, memorable, single-sitting text — well suited to small-group and discipleship use.
• Explicit teaching notes on the ἐλλογάω/imputed-righteousness cross-reference and the mui-tsai/外傭 distinction turn documented risks into strong discipleship teaching moments rather than merely defensive translation constraints.

Recommended actions

1. Route Philemon 1:8-21 to human theologian review as a single unit, not per-segment, per the escalation rule in 12_ai_translation_requirements.md.
2. Enforce the extended assets/translation_memory.json (59 terms + 11 proper nouns) with zero deviation on all 16 Critical/High terms (6 Critical + 10 High) — 12 of the 21 doctrines require mandatory human theologian review; 6 require native speaker review; 3 are automated-only.
3. Apply the mandatory [TRANSLATOR NOTE] triggers at every 奴僕, 弟兄 (1:16), ἐλλογάω (1:18), house-church (1:2), πρεσβύτης variant (1:9), Onesimus pun (1:11, 1:20), and human-spirit (1:25) occurrence before any segment is marked approved.
4. Adopt YouVersion CUV (version ID 46, book code PHM) as the default hyperlink target and 腓利門書 as the fixed citation book name, consistent with 05_translation_landscape.md.
5. Validate σπλάγχνα consistency (1:7, 1:12, 1:20) as its own Phase 2 Step 15 quality-score dimension, given its documented inconsistency in existing Chinese Bible translation traditions.