CATTI Level 2 cheat sheet
CATTI Level 2 · Keep the key ideas within reach.
笔译 (Translation) - CATTI Level 2 Reference Sheet
笔译实务 (Level 2 practical translation) is a handwritten/typed test of two passages: 英译汉 (E→C) and 汉译英 (C→E), each roughly 300–400 source words/characters, with a total time limit of 180 minutes; 综合能力 (comprehensive ability) is a separate 120-minute objective paper (vocabulary, grammar, reading, cloze).
Scoring emphasis in 实务 is accuracy of meaning first, then target-language fluency; a mistranslated key term or reversed logic (negation, causality, condition) is a 重大错译 and costs far more than an awkward but faithful rendering.
Decision rule for long English sentences: identify the main clause (S+V+O) first, then map subordinate clauses and modifiers; in E→C recast into 流水句 (short sequential clauses) rather than mirroring the English tree structure.
Decision rule for Chinese 四字格 and 排比: do not translate character-by-character; render the underlying proposition, and drop or merge parallel items only when the repetition carries no distinct information.
Terminology handling: proper nouns, organization names, and technical terms follow established official renderings (e.g. UN, WTO, IMF terminology databases); where no fixed equivalent exists, translate meaning plus a brief in-text gloss rather than transliterating blindly.
Numbers, units, dates, and negation are high-risk: 亿 = 100 million, 万亿 = trillion; check 翻番 (double), 增长/增长了 (increase by) vs 增长到 (increase to), and 不/无/未/非 scope before finalizing.
Cohesive devices differ by direction: English→Chinese typically deletes or varies pronouns and uses 意合 (implicit linkage); Chinese→English typically adds explicit connectives (however, therefore, while) and subjects that Chinese leaves understood.
Register and genre control: formal documents (government work reports, treaties, contracts) require consistent term choice across the whole passage and a formal C→E register; avoid colloquial contractions and avoid over-literal calques such as 'strengthen the construction of'.
Time budget for 实务: about 80–85 minutes per passage plus 10–15 minutes for review; review checklist = completeness (no dropped sentence), figures/dates, proper names, negation, and target-language grammar agreement.
口译 (Interpreting) - CATTI Level 2 Reference Sheet
Level 2 口译实务 is 交替传译 (consecutive): 英译汉 and 汉译英 segments of roughly 1–2 minutes each, delivered after the speaker stops; 同声传译 (simultaneous) is a separate Level 2 subject with near-real-time delivery and a lag of about one clause.
Note-taking in consecutive uses symbols and abbreviations, not shorthand: record propositions (subject–verb–object), numbers, proper nouns, and logical markers (because, but, therefore); notes are retrieval cues, and the target must not be read verbatim from the page.
Effort management in simultaneous interpreting follows Gile's Effort Model: listening/analysis + short-term memory + production + coordination must stay under total capacity; when a segment is dense, prioritize propositions over adjectives and stylistic ornament.
Numbers, units, and rates are the top error source: convert 亿/万亿 to million/billion instantly (亿 ×100 = million; 万亿 = trillion), and pre-prepare percentage, growth, and ranking formulas before the test.
Anticipation and reformulation rules: predict likely collocations from context (e.g. 'on behalf of… I would like to…'), and when a source structure has no target equivalent, reformulate the meaning rather than produce a word-for-word calque that breaks target grammar.
Handling unknown terms: keep the original term with a generic classifier (e.g. 'the so-called X project'), or describe function; never invent a plausible-sounding technical equivalent, and never stop the delivery to explain.
Delivery criteria: fluency (limited filled pauses and self-repair), complete proposition coverage, correct names and figures, and appropriate register; omissions of whole propositions and number errors are the heaviest deductions.
Self-monitoring rule: if an error is caught within a few seconds, correct it briefly and continue; if not, do not retro-correct mid-segment - a long backtrack costs more than the original slip.
Preparation routine for the test room: confirm audio level and microphone, keep note paper oriented, and use the first seconds of each segment to lock the topic, speaker stance, and tense before writing.