Reading: Paper 1 Section B, and how to read Chinese without a dictionary

What Section B is, and what its texts are actually like#

StandardPearson Edexcel 9CN0 Paper 1 Section B - reading, 30 marks, recommended 50 minutes, four questionsDfE GCE AS and A level subject content for modern foreign languages

Section B carries 30 of Paper 1's 80 marks, the same as listening, and therefore 15 per cent of the whole A-Level. It presents four questions on Chinese texts, using both multiple-choice and open-response formats, and every question is set in English and must be answered in English. The specification describes the texts as authentic or adapted from authentic sources, containing both factual and abstract content, written for different purposes and audiences, and drawn from the four themes. Read that description carefully: different purposes and audiences means the register will vary within the paper, and abstract content means at least one text will be an argument rather than a report. The phrase different purposes and audiences is worth pausing on, because it tells you the questions will vary in kind as well as in difficulty. A public notice invites a factual question; an opinion column invites an inference question; a personal essay invites a question about attitude. Recognising the text type therefore predicts the question type before you have read a word of content.

The consequence for preparation is that a diet of textbook passages is not enough. A textbook passage is written to be understood by a learner; an authentic passage is written to inform or persuade a Chinese reader and assumes shared knowledge you may not have. The remedy is not to read harder material but to read a wider spread of it: a news report, an opinion column, a personal essay or blog post, a public-information notice, a book or film review. Each has its own conventions, and meeting a convention for the first time under examination conditions costs you comprehension that has nothing to do with your Chinese. The practical version of that advice is to read in Chinese for fifteen minutes a day rather than for two hours a week, and to vary the source rather than the length. A news app, a review site, a public-information page and somebody's blog in a single week will teach you more about register than a term of graded readers.

The single most important fact about Section B is that no dictionary is permitted. Every candidate will meet characters they cannot read and words they do not know. The examination is therefore not testing whether you know every word; it is testing whether you can navigate a text that contains unknown words and still answer four precise questions. That reframing matters psychologically as well as strategically: encountering an unknown compound is the normal condition of reading in this examination, not a sign that something has gone wrong. There is a psychological corollary that matters more than it sounds. Candidates who expect to understand everything panic at the first unknown compound and read the rest of the passage in a degraded state; candidates who expect gaps navigate around them and read on. The expectation itself is worth adjusting deliberately, because it costs nothing and it protects the second half of every text.

Finally, the answering. Because you answer in English, the marks are for comprehension and for precise expression, not for Chinese production. That means every answer should be short, exact and matched to the interrogative in its stem; that a translated clause is not the same as an answer; and that where a question carries three marks it is telling you it wants three distinct points. Nothing you write in Section B is assessed for its Chinese, which is worth remembering when you are tempted to hedge an English answer with a Chinese term you are unsure how to render. One more consequence: your English matters. An answer that renders 部分 as most, or 可能 as certainly, or 有些人 as people, has misreported the Chinese even though the reading was correct. Precision in English is therefore part of the skill being assessed, and it is worth checking your practice answers against the strength of the original rather than only against its content.

Worked example

Identifying purpose and audience before reading for detail

Four openings, one from each of four text types. Say in each case who the text is for, what it is trying to do, and what that predicts about the questions. (1) 据市教育局介绍, 从下学期起, 全市中小学将统一推迟早上上课时间. (2) 说实话, 我一开始并不喜欢这部电影. (3) 每年春节前后, 铁路部门都会面临巨大的压力. (4) 请各位乘客注意: 本站出口临时关闭, 请从东门离开.

Text 1 - the source formula tells you everything
据…介绍 (according to information from…) plus a named official body plus 将 (will, formal futurity) plus 统一 (uniformly) marks this as news reporting of an official announcement. Audience: general readers. Purpose: to inform. Prediction: questions will ask for stated facts - what changes, from when, who is affected - and possibly for a stated reason. Inference questions are less likely here.
Text 2 - the personal marker changes the question type
说实话 (to be honest) plus 我 plus 并不喜欢 (did not at all like) plus 一开始 (at first) marks a personal review. 一开始 is the giveaway: it promises a change of mind, so the argument of the piece is a reversal. Audience: readers deciding what to watch. Purpose: to evaluate. Prediction: questions about the writer's opinion and how it changed - inference rather than fact.
Text 3 - the general-truth opening promises an explanation
每年…都会 (every year… invariably) plus 面临 (to face) plus 巨大的压力 marks expository writing that states a recurring phenomenon in order to explain it. Audience: general readers. Purpose: to explain. Prediction: the questions will follow the causal chain, and there will almost certainly be a why question worth three marks.
Text 4 - the formal-notice register
请各位乘客注意 plus the colon plus 本站 (this station, with the formal 本) plus 临时 (temporarily) marks a public notice. Audience: people physically present. Purpose: to instruct. Prediction: a short factual question, possibly multiple choice, about what a reader should do. Note 本 here: it is one of the Appendix 3 formal pronouns and its presence alone signals institutional register.
Why the two minutes are worth spending
You now know, before reading a single content sentence, whether to expect fact questions or inference questions, whether the argument reverses, and where the causal chain will be. That is not guesswork - it is reading the conventions of the text type, and the conventions are as reliable in Chinese as they are in English. It also tells you which text to do first: informational ones are usually quicker, so bank them and give the argumentative text the time it needs.
What was actually assessed
Nothing here required advanced vocabulary. 据…介绍, 说实话, 每年…都会 and 请…注意 are all short, high-frequency formulae, and each of them is worth more to a reader than ten content words, because each one predicts the shape of what follows. Collect these openings as you read, exactly as you would collect vocabulary.

Result: (1) Official news report, general audience, informing: expect stated-fact questions. (2) Personal review, prospective viewers, evaluating: 一开始 promises a reversal, so expect opinion and inference questions. (3) Expository explanation, general audience: expect a causal chain and a three-mark why question. (4) Public notice, people present, instructing: expect one short factual question, and note 本站 as a formal-register marker.

Typical mistakes

Preparing only with graded readers. The specification says authentic or adapted from authentic sources.
Answering in Chinese, or leaving a Chinese term untranslated in an English answer.
Treating an unknown word as a stopping point. It is the ordinary condition of the task.
Giving one point for a three-mark question. The tariff is the instruction.

Active revision

找五篇不同类型的中文文章 (新闻报道, 评论, 个人随笔, 公告, 影评), 每篇大约三百字, 每天读一篇, 第一遍不查任何词. 读完以后用英文写三句话: 这篇文章是写给谁的, 目的是什么, 主要论点是什么.

Active recall

Recall the key points — then reveal.

Reading without a dictionary: getting meaning out of unknown characters#

AdvancedPearson Edexcel 9CN0 Paper 1 Section B - reading; no dictionaries permittedDfE GCE AS and A level subject content for modern foreign languages

Chinese offers a reader three independent routes to the meaning of an unknown word, and using them in order is a trainable procedure. Route one is morphology: most modern Chinese words are compounds of two morphemes you probably know. 环保 is 环境 plus 保护, environmental protection. 高铁 is 高速 plus 铁路, high-speed rail. 网购 is 网络 plus 购买, online shopping. 留守儿童 is stay-behind plus children. When you meet an unfamiliar two-character word, split it and ask what each half means elsewhere; you will be right far more often than you expect. The morphological route is unusually productive in Chinese because the modern written language was largely built by compounding, often deliberately, during the vocabulary expansion of the last century and a half. That means the transparency is not accidental: a great many technical and abstract terms were assembled from morphemes chosen precisely so that a literate reader could parse them.

Route two is the character itself. Most characters contain a semantic component that narrows the field: 氵 for water and liquids (河, 江, 汽, 污), 讠 for speech (说, 语, 读, 论), 忄 for feelings (情, 快, 慢, 怕), 木 for wood and plants, 钅 for metal, 疒 for illness, 辶 for movement. This will rarely give you the word, but it will often give you the domain, and the domain plus the sentence is frequently enough. 污染 with 氵 twice, in a sentence about factories and air, is not going to be anything other than pollution. A short list of high-yield components is worth learning outright, because they recur constantly in the four themes: 讠 speech, 氵 water, 忄 feeling, 辶 movement, 钅 metal, 木 wood, 疒 illness, 扌 hand and action, 女 female, 亻 person, 宀 building, 艹 plants, 火 fire, 目 eye, 心 heart, 贝 money and value.

Route three is the sentence's own grammar, and it is the most powerful of the three. Chinese marks structure clearly: 的 tells you a modifier just ended and a noun follows; 地 tells you a manner adverb precedes a verb; 得 tells you a complement follows a verb; 了, 过 and 着 tell you the preceding item is a verb; 把 and 被 tell you what follows is a verb phrase acting on a specified object; 是…的 brackets an emphasis. An unknown item sandwiched between 很 and 的 is an adjective. You very often do not need to know what a word means in order to know what it is doing, and knowing what it is doing is frequently enough to answer the question. There is a fourth route worth adding for abstract prose: parallelism. Chinese argument is frequently built in balanced pairs, so an unknown item in one half of a pair is constrained by the known item in the other. 不是…而是…, 一方面…另一方面…, 有的…另一些… all guarantee that the two halves are comparable, which often narrows an unknown word to a near-opposite of one you know.

Three practical rules complete the method. First, read the whole sentence before deciding an unknown word matters; a great many unknown words are elaborations that the question does not touch. Second, if an unknown word appears three times it is probably the topic, and it is worth spending thirty seconds triangulating it from its three contexts. Third, if it appears once and is not in the question, ignore it entirely and move on - the commonest way to lose Section B marks is to spend four minutes on a word that carried none. Practise the procedure with a stopwatch, because its value is as much about speed as about accuracy. Fifteen seconds per unknown word is a working budget; anything longer and you should be moving on and marking it. Candidates who have never timed themselves consistently overspend on the first unknown word in a passage and underspend on everything afterwards.

Worked example

Four unknown words in one paragraph, solved three ways

Read this and work out the four marked items without a dictionary. 随着城市不断扩张, 郊区的农田正在迅速减少. 有的地方把原来的村庄改造成了商业区, 居民只能搬迁到更远的地方. 环保组织对此提出了质疑, 认为规划部门低估了生态影响. Assume you do not know 扩张, 搬迁, 质疑 or 低估.

扩张 - morphology plus context
Split it: 扩 appears in 扩大 (to enlarge) and 张 in 张开 (to open out, to spread). Both point to expansion. The context confirms it: 随着城市不断扩张 sits in a sentence whose second half says farmland is rapidly decreasing. A city doing something continuously that causes farmland to shrink is expanding. Decision: expansion, and the grammar tells you it is a verb or verbal noun following 城市.
搬迁 - morphology, and the sentence's own logic
搬 you know from 搬家, to move house. 迁 is less familiar but shares its semantic field. The structure settles it anyway: 居民只能搬迁到更远的地方 - residents can only 搬迁 to a more distant place. A verb taking 到 plus a place, with 居民 as its subject and the previous sentence describing villages being turned into commercial districts, can only mean relocate. Decision: to be relocated, to move away.
质疑 - the semantic component does the work
Both characters carry 讠, the speech radical - so this is something done with words. 疑 appears in 怀疑, to doubt. The structure: 环保组织对此提出了质疑 - an environmental organisation 提出了 (put forward) a 质疑 about this. A noun put forward by a critical organisation, built from speech plus doubt, is a challenge or a questioning. Decision: raised objections, called it into question.
低估 - morphology alone, but check the direction
低 is low and 估 appears in 估计, to estimate. So: to estimate low, to underestimate. Check the direction against the argument, because getting it backwards would reverse the whole sentence: the environmental group is objecting, so it must be alleging that the planners got the impact wrong in the direction that favours the development - underestimated, not overestimated. Decision: underestimated. Note that the opposite, 高估, exists and works the same way.
Now confirm you can answer without perfect knowledge
Suppose 质疑 had defeated you entirely. You would still know from 环保组织对此提出了…, 认为规划部门低估了生态影响 that the environmental organisation objected and thought the planners had underestimated the ecological impact - because 认为 introduces their position directly. An unknown noun is often bypassed by the clause that explains it, and looking for that clause is faster than solving the word.
What was actually assessed
Four unknown words, four different routes: morphology, structure plus logic, semantic component, morphology with a direction check. None required a dictionary and none took more than fifteen seconds once the procedure was habitual. Build the habit by doing this deliberately on every text you read outside the exam, writing your guess down before checking, so that you know your own hit rate.

Result: 扩张 = to expand (morphology: 扩大 plus 张开, confirmed by farmland decreasing). 搬迁 = to relocate (搬家 plus a verb taking 到 plus a place, with residents as subject). 质疑 = to call into question (both characters carry the speech radical; 疑 from 怀疑, put forward by a critical body). 低估 = to underestimate (低 plus 估计, direction checked against the objection being made).

Typical mistakes

Stopping at the first unknown character. Read to the end of the sentence first; the sentence usually supplies the constraint.
Guessing from the phonetic component. It is unreliable for a learner and the semantic component is far safer.
Assuming a compound means the sum of its parts literally. 打交道 is not to hit an intersection; use morphology as a hypothesis, then test it against the sentence.
Spending disproportionate time on a word that no question asks about. Time is the scarce resource in Section B, not vocabulary.

Active revision

拿一篇你没读过的中文文章, 圈出所有不认识的词. 对每一个词写下三件事: 拆开以后两个语素分别是什么意思; 里面有没有你认识的偏旁; 它在句子里是名词, 动词还是形容词 (根据的, 地, 得, 了, 很判断). 猜完以后再查字典, 统计你的正确率.

Active recall

Recall the key points — then reveal.

The skeleton of a Chinese paragraph: connectives, topic chains and formal register#

AdvancedPearson Edexcel 9CN0 Appendix 3 - conjunctions, formal pronouns, structural markersPearson Edexcel 9CN0 Paper 1 Section B - readingDfE GCE AS and A level subject content for modern foreign languages

An argumentative Chinese paragraph is held together by a small set of connectives, and reading them accurately gives you the argument even where the content defeats you. Addition and escalation: 不但…而且…, 不仅…还…, 甚至. Contrast: 但是, 可是, 然而 (formal), 不过, 却 and 反而 (both pre-verbal, both signalling something contrary to expectation), 相反. Cause and consequence: 因为…所以…, 由于, 因此, 从而, 之所以…是因为…. Concession: 虽然…但是…, 尽管…仍然…, 就算…也…. Condition: 如果…就…, 只要…就…, 只有…才…, 除非. Deduction: 由此可见, 可见, 换句话说. Learn them as a reading set, in these groups, and half of Section B's inference questions become mechanical. It is worth learning the register variants alongside the plain ones, because Section B texts are drawn from authentic sources and the formal member of a pair is what you will actually meet: 然而 for 但是, 由于 for 因为, 因此 for 所以, 此外 for 还有, 尽管 for 虽然, 并且 for 而且. Reading them as unfamiliar words rather than as known ones is a wholly avoidable difficulty.

Chinese also organises paragraphs with topic chains, which look like missing subjects to an English reader. Once a topic is established, subsequent clauses simply omit it: 小王昨天去了图书馆, 借了三本书, 一直看到晚上才回家. Three clauses, one subject, stated once. The reading rule is that an apparently subjectless clause almost always continues the previous topic - and the moment a new noun appears in subject position, the chain has been broken deliberately, which is itself information. English translations must restore the subject each time, which is why a Section C translation of such a passage is longer than its original. Topic chains also explain a phenomenon that puzzles learners: Chinese paragraphs often look shorter than their English translations. Every clause that would carry a pronoun in English carries nothing in Chinese, and the saving accumulates. When translating in Section C, expect your English to be longer than the source, and do not assume you have added something.

Register is marked by a handful of items that a learner can spot in seconds. The formal pronouns and demonstratives from Appendix 3: 该 (that aforementioned, as in 该政策), 此 (this, as in 与此同时, 对此), 本 (this present, as in 本站, 本地), 其 (its, their, as in 其中, 其他). The classical possessive 之 in fixed expressions: 五分之三, 之一, 总之. Nominalising 所: 所说的, 所需要的. Bound morphemes: 性 for abstract nouns (灵活性, 可能性), 化 for processes (城市化, 商业化), 者 for agents (作者, 支持者, 消费者). Any of these in the first two lines tells you the text is expository rather than conversational, and expository texts reward reading the connectives first. Two of these deserve extra attention because they are structurally load-bearing rather than merely formal. 其中 picks one item out of a set just mentioned, so it always points backwards and tells you the previous sentence contained a group. And 所 nominalises the verb that follows it, so 所说的 is what was said and 所需要的 is what is needed - a construction that turns a clause into a noun and is very common in expository writing.

Put the three together and you have a reading procedure for a hard paragraph. First scan for connectives and mark them; they give you the argument's shape before you know its content. Second, identify the topic and assume it continues until a new subject appears. Third, note the register markers and adjust your expectations of the vocabulary. Only then read for content. Candidates who read a hard Chinese paragraph linearly from the first character usually understand the first sentence and lose the argument; candidates who read the skeleton first usually get the argument and can afford to be vague about several sentences. The procedure has one further benefit worth naming. Scanning for the skeleton first tells you how many claims the paragraph makes, and therefore how many marks a question on it is likely to carry. A paragraph with a concession, a deduction and a conditional is a three-mark paragraph; one with a single 因为 is not, whatever its length.

Worked example

Reading a dense paragraph by its skeleton first

尽管近年来不少城市都推行了垃圾分类, 效果却不完全一致. 有的地方居民参与度很高, 分类准确率也在稳步上升; 另一些地方则流于形式, 垃圾在收运环节又被重新混在一起. 由此可见, 政策本身并不足以改变行为, 关键还在于后端的处理能力. 换句话说, 如果分好的垃圾最后还是被一起处理, 那么再严格的规定也很难长期维持居民的积极性.

Pass one: mark the connectives only, read no content
尽管…却…; 有的…另一些…则…; 由此可见…; 换句话说…; 如果…那么…也很难…. That is five structural markers. Before knowing a single fact, you can say: a concession, then a two-way contrast between cases, then a deduction, then a restatement, then a conditional conclusion. The paragraph is an argument in five moves and you have its shape.
尽管…却… - the concession sets the thesis
尽管 concedes that something is true, 却 (pre-verbal) delivers the contrary point. So: although X has happened, nevertheless Y. The thesis of the paragraph is Y, not X. Reading only these two words tells you that whatever comes after 却 is what the writer wants you to take away - 效果却不完全一致, the results have not been uniform.
有的…另一些…则… - the evidence is a contrast of cases
则 is the formal correlative adverb marking the second case against the first. So the two clauses are not a list but a paired comparison, and any question asking why results differ will be answered from this pair. 流于形式 (to become a mere formality) is the phrase carrying the negative case, and 又被重新混在一起 - are mixed together again - is the mechanism.
由此可见 - the deduction, which is the answer to the main question
由此可见 (from this it can be seen) announces the paragraph's conclusion, and in an examination this is where a three-mark why or what does the writer conclude question is answered. 政策本身并不足以改变行为 - policy alone is not enough to change behaviour - and 关键还在于 (the key lies rather in) points to the real factor: 后端的处理能力, downstream processing capacity.
换句话说 plus 如果…那么… - the restatement makes it concrete
换句话说 tells you the next sentence repeats the conclusion in other words, so if you understood the previous sentence you can read this one fast, and if you did not, this is your second chance. The conditional then supplies the mechanism in concrete form: if sorted waste ends up processed together anyway, no regulation however strict can sustain residents' willingness.
What was actually assessed
The paragraph contains at least four words a candidate might not know - 参与度, 稳步, 流于形式, 收运 - and none of them was needed. The argument was carried entirely by 尽管…却…, 则, 由此可见, 换句话说 and 如果…那么…. That is the point of this section: in expository Chinese the connectives are the argument, and a reader who owns thirty of them can follow a text whose vocabulary is well beyond them.

Result: Structure: concession (many cities have introduced waste sorting) - but (results have not been uniform) - paired evidence (some places have high participation and rising accuracy; others let it become a formality, and the waste is remixed at collection) - deduction (policy alone cannot change behaviour; the key is downstream processing capacity) - restatement as a conditional (if sorted waste is processed together anyway, no regulation however strict can sustain residents' willingness for long).

Typical mistakes

Treating 却 and 反而 as ordinary conjunctions. Both are pre-verbal adverbs and both reverse an expectation, which is usually the point of the sentence.
Reading a subjectless clause as impersonal. It continues the topic, and mis-assigning it can invert an answer.
Ignoring 其中. It means among them and it is constantly used to pick out one item from a set just mentioned.
Reading linearly through a dense expository paragraph. Scan the connectives first; the shape makes the content easier.

Active revision

拿一段中文议论文, 先只圈出所有连接词和该, 此, 本, 其, 所, 性, 化, 者这些标记, 不读内容. 只根据这些标记用英文写出这段话的论证结构 (先说什么, 然后让步什么, 最后推出什么). 再完整读一遍, 看你的结构猜对了几成.

Active recall

Recall the key points — then reveal.

Answering in English: precision, tariff and the interrogative#

StandardPearson Edexcel 9CN0 Paper 1 Section B - questions set in English, answered in EnglishDfE GCE AS and A level subject content for modern foreign languages

Because the answers are in English, Section B quietly assesses two things at once: whether you understood the Chinese, and whether your English says exactly what you understood. The second is where careful candidates lose marks they had already earned. An answer that says the writer thinks pollution is bad when the text said the writer thinks the response to pollution has been uneven is an English failure sitting on top of a successful reading. Aim for answers that are short, literal about the claim, and faithful to its strength: 明显 is marked, not huge; 有些人 is some people, not many; 可能 is may, not will. The general rule is that your English should be no more confident than the Chinese was. Chinese hedges with 可能, 也许, 未必, 大概, 一般来说 and 有些 quite as carefully as English does, and every one of those words has a job. Reproducing the hedge is not timidity; it is accuracy, and dropping it converts a careful claim into an overstatement the text never made.

The tariff is an instruction. One mark means one point and nothing more; adding a second costs time and can introduce an error. Three marks means three distinct points, and the marker needs to see them as three - so separate them with semicolons or number them. If you cannot find three, give the two you are sure of and one plausible third; there is no penalty for a wrong point in an open-response item of this kind, and there is a certain loss for a missing one. The reflex to build is: read the tariff, count the points, then write. There is one qualification to the no-penalty rule worth stating. A third point that contradicts your other two suggests you have not understood the passage, and although it will not be deducted, it dilutes an answer a marker is reading quickly. Give a plausible third point, not a desperate one, and make sure it is consistent with the two you are confident of.

Match the first words of your answer to the interrogative. Why becomes Because…. How becomes By…. In what way becomes By doing X rather than Y. What evidence becomes The evidence is that…. What does the writer suggest becomes The writer suggests that…. This sounds mechanical and it is, and it is the most reliable single habit in the section, because the commonest near-miss is an answer that is entirely true and does not answer the question that was asked. If your answer does not begin with the interrogative's own shape, check that it is answering rather than reporting. The habit is easy to test on your own past work. Take five of your answers, cover the questions, and try to reconstruct the interrogative from the answer alone. Where you cannot tell whether the question asked why or what, the answer was not shaped by its stem - and that shapelessness is exactly what a marker sees when deciding whether the point was made.

Two more precision points. First, attribution: many Section B texts contain more than one voice - the writer, a quoted expert, an anonymous 有人. A question that says according to the researcher wants the researcher's view, and giving the writer's instead scores nothing even when the content is well expressed. Underline the attribution phrases as you read: 一位专家指出, 有人认为, 报道称, 作者本人则. Second, do not lift a translated clause verbatim as an answer where the question implies your own words; a marker cannot tell whether you understood it or matched a string of characters. One more precision point that costs nothing: keep tense and modality faithful. A Chinese text that says 政府提出了一个目标 says the target was put forward, not that it was met; a text that says 可能会改变 says it may change, not that it will. English tense and modal verbs carry distinctions Chinese carries lexically, and a careless English verb can silently invent a fact.

Worked example

Four answers, three of them nearly right

Text extract: 一位社会学者指出, 年轻人回到二线城市, 主要不是因为大城市的机会变少了, 而是因为生活成本上升得太快. 报道还提到, 有些受访者表示, 家人也是一个重要因素. 作者本人则认为, 这两种解释并不矛盾. Questions: (a) According to the sociologist, why are young people returning to second-tier cities? [2] (b) What additional factor did some interviewees mention? [1] (c) What is the author's own view? [1]

(a) The near-miss: reporting instead of answering the contrast
A weak answer: Because there are fewer opportunities in the big cities. That is what the text explicitly DENIES - 主要不是因为…, 而是因为…. The 不是…而是… frame is the whole content of the sentence, and an answer that takes the rejected half has inverted it. Two marks are almost certainly one for the rejection and one for the real cause.
(a) The answer that earns both marks
Because, in his view, the main cause is not a decline in opportunities in the big cities but the fact that the cost of living there has risen too quickly. Note that it begins with Because, preserves both halves of 不是…而是…, keeps 主要 as the main cause rather than dropping it, and renders 太快 as too quickly rather than as quickly - the 太 is a judgement and it belongs in the English.
(b) One mark, one point, and the right voice
有些受访者表示 - some interviewees said. The answer is family. But note the voice: this is not the sociologist and not the author, and a candidate who has not been tracking attributions may hand it to the wrong one. Answer: Family, which some of those interviewed said was also an important factor. Do not add anything; the tariff is one.
(c) The trap of the obvious
作者本人则认为, 这两种解释并不矛盾. 则 marks the author's position against the preceding ones, 本人 emphasises that it is his own, and 并不矛盾 is emphatic negation - not at all contradictory. A weak answer says the author agrees with the sociologist, which is not what he says. Answer: That the two explanations are not in conflict with each other.
Track the voices as a separate pass
This short extract contains three voices in four sentences: 一位社会学者, 有些受访者 and 作者本人. All three questions turn on getting the voice right, and none of them turns on difficult vocabulary. When a text introduces a second voice, mark it in the margin; it is a reliable predictor of where the questions will be.
What was actually assessed
Comprehension of 不是…而是…, of 主要, of 太, of 则 and 本人, and of 并不 - six structural items, all of which change an answer if missed. Plus English precise enough to keep main, too quickly and not at all. Nothing in Section B rewards elegant English; everything rewards exact English.

Result: (a) Because, in his view, the main reason is not that opportunities in the big cities have declined but that the cost of living there has risen too quickly. (b) Family: some of the interviewees said this was also an important factor. (c) That the two explanations are not contradictory.

Typical mistakes

Overstating a hedged claim. 可能 rendered as will is a comprehension error dressed as an English one.
Writing a paragraph for a one-mark question. It gains nothing and costs the time a three-mark question needs.
Translating a clause and calling it an answer. Translation reports; an answer explains, selects or infers.
Attributing to the writer what a quoted expert said. The stem usually specifies the voice.

Active revision

拿五道你做过的阅读题, 只看你的答案, 不看原文, 判断每个答案回答的是哪个疑问词 (why, how, what evidence, who thinks). 然后对照题干, 找出答非所问的那几个, 重写, 每个答案都以疑问词对应的英文开头.

Active recall

Recall the key points — then reveal.

Exam technique: question-type triage and the fifty minutes#

StandardPearson Edexcel 9CN0 Paper 1 Section B - 30 marks, four questions, recommended 50 minutesDfE GCE AS and A level subject content for modern foreign languages

Four questions, 30 marks, 50 recommended minutes - about twelve minutes per question with two minutes spare. But the marks will not be spread evenly, and the first act of the section should be to see how they are spread. Read all four questions and their tariffs before reading any text, and write a target time beside each. A question worth 10 marks deserves roughly a third of the section; a multiple-choice item worth 2 does not deserve five minutes however tempting the distractors are. Candidates who do not do this arithmetic almost always over-invest in whichever question they happened to find hardest. There is one further piece of arithmetic worth doing at this point: how many texts there are. Four questions may sit on one text or on two or three, and if two questions share a text they should be answered together, because the close read is the expensive part and doing it twice is pure waste. Note which questions share a text before you begin.

Then triage by type rather than by order. Do the factual and multiple-choice items first: they are quick, they bank marks, and doing them familiarises you with the text so that the harder questions are easier when you reach them. Leave inference and summary items for the second half, when you have read the text twice by accident. The only exception is a very high-tariff item, which should be started while you are freshest even if it is the hardest - not finished, but started, so that you are not composing its answer at minute forty-nine. The exception has its own exception worth naming. If the highest-tariff question is a summary spanning two texts, it cannot be started until both have been read - so in that case the correct move is to skim both early and then bank the factual questions while the shape of both texts is fresh, rather than starting the summary out of order.

Read the questions before the text. This is worth stating on its own because it feels wrong and is right. Reading the questions first turns reading the text from comprehension into a search: you know what you are looking for and can afford to skim what you are not. It also protects you from the commonest time sink in Section B, which is decoding a paragraph in full that no question touches. Skim first, then read closely only the parts the questions point at - and remember that abstract, argumentative paragraphs are usually where the high-tariff questions are. The saving is measurable if you want to convince yourself. Time two practice sections, one reading text-first and one reading questions-first, and compare both the time taken and the number of near-misses. Almost everybody finds the second faster, and the reason is not speed of reading but that a search terminates when it succeeds while a comprehension read does not.

Reserve the last five minutes for a specific check, and make it the same check every time: read each stem and each answer against it, in order, asking only whether the answer answers the question. Not whether it is true, not whether the English is good - whether it answers. That check finds the near-misses, and near-misses are where the difference between a good and an excellent Section B mark lives. If you find you have time beyond that, spend it adding a point to the highest-tariff open-response question rather than polishing prose that is already adequate. It is worth deciding in advance what to do if you finish early, because the instinct to reread the Chinese is strong and unproductive at that stage. The two useful actions are adding a point to the highest-tariff open-response question and checking that multi-mark answers are visibly separated. Both add marks; rereading the passage almost never does.

Worked example

A 50-minute Section B, planned and defended

The section has four questions on two texts. Q1 multiple choice on text A, 4 marks. Q2 detail on text A, 6 marks. Q3 inference on text B, 8 marks. Q4 summary across both texts, 12 marks. Build the plan and say what to do when it slips.

Minutes 0 to 4: read all four stems and do the arithmetic
Note that Q4 carries 12 of the 30 marks - 40 per cent of the section - and that it spans both texts, which means it cannot be attempted until both have been read. Write targets: Q1 6 min, Q2 9 min, Q3 12 min, Q4 18 min, check 5 min. That sums to 50. Writing the sums down is what stops you spending fourteen minutes on Q3.
Minutes 4 to 10: skim both texts for shape, not content
Purpose, audience, register, and the position of the connectives. You are not trying to understand yet. At the end of six minutes you should be able to say what each text is doing and roughly where in each the answers to Q2 and Q3 live. This makes every subsequent read a targeted one.
Minutes 10 to 25: Q1 then Q2, banking the factual marks
The multiple choice and the detail question are both on text A and both largely factual, so doing them together means reading text A closely once and answering twice. This is the practical reason for triaging by type rather than by order: it minimises the number of close reads.
Minutes 25 to 37: Q3, the inference question
Text B, read closely, connectives first. Inference questions are answered by the connective tissue - 其实, 却, 反而, 并不, 未必 - far more often than by content words, so a first pass marking those is not a detour. Write the answer immediately while the reasoning is fresh; inference answers deteriorate if left.
Minutes 37 to 45: Q4, the summary across both texts
Eight minutes for twelve marks is tight, which is why the skim at minutes 4 to 10 mattered - you already know the shape of both texts and where they agree and differ. Twelve marks suggests six or more distinct points; list them in note form first, then write. A summary written straight into prose loses points to fluency.
Minutes 45 to 50: the answering check
Stem by stem, in order. Does Q2 give six marks worth of distinct points? Does Q3 begin with the shape of its interrogative? Does Q4 attribute where the two texts differ? Nothing else. Do not re-read the Chinese; the errors you can still fix at this point are answer-shaped.
What was actually assessed
When it slips: if Q3 overruns, cut it at its target time and move to Q4 regardless, because twelve marks unstarted beats eight marks perfected. If the skim takes ten minutes rather than six, take the four minutes from Q1, which is the cheapest question to rush. Decide these trade-offs now; at minute forty-three under exam conditions you will not weigh them well.

Result: 0-4 read all stems, write time targets from the tariffs; 4-10 skim both texts for purpose, register and connective positions; 10-25 answer Q1 and Q2 together on one close read of text A; 25-37 read text B connectives-first and answer Q3 immediately; 37-45 list points then write the 12-mark cross-text summary; 45-50 check each answer against its stem for answering only. If a question overruns, cut it at its target and protect the highest-tariff question.

Typical mistakes

Reading the whole text closely before looking at the questions. It is the single largest avoidable time loss in the section.
Working through the questions in the printed order regardless of tariff and type.
Spending disproportionate time on a hard multiple-choice item. Its mark is the same as any other.
Using the final minutes to improve English style rather than to check that each answer answers.

Active revision

做一次完整的五十分钟模拟: 先读四道题和分数, 写下每题的目标时间; 再快速浏览文章; 先做事实题和选择题, 再做推断题和总结题; 最后五分钟只做一件事, 逐题对照疑问词检查. 记录实际用时和计划的差距.

Active recall

Recall the key points — then reveal.

Key points

Identifying purpose and audience before reading for detail

Key points

Four unknown words in one paragraph, solved three ways

Key points

Reading a dense paragraph by its skeleton first

Key points

Four answers, three of them nearly right

Key points

A 50-minute Section B, planned and defended

Reading: Paper 1 Section B, and how to read Chinese without a dictionary

Key points

Identifying purpose and audience before reading for detail

Key points

Four unknown words in one paragraph, solved three ways

Key points

Reading a dense paragraph by its skeleton first

Key points

Four answers, three of them nearly right

Key points

A 50-minute Section B, planned and defended