EuraStudy
Notes/Japanese/Reading: Paper 1 Section B, and the reading half of Paper 3 Section B
Notes · JapaneseUK · A-Levels

Reading: Paper 1 Section B, and the reading half of Paper 3 Section B

This topic carries the two places where reading is assessed on its own: Paper 1 Section B, which is 20 marks in a recommended 45 minutes across four questions, and the reading half of Paper 3 Section B, where a written source is summarised at 5(b) and then weighed against a heard source at 5(c). The two are marked by the same assessment objective, AO2, but they differ in a way that is easy to get wrong and expensive to get wrong: Paper 1 Section B's questions are set in English and are answered in English, while everything in Paper 3 Section B is set in Japanese and answered in Japanese. Around that structural fact the topic teaches what the specification says about the texts themselves — authentic or adapted, factual and abstract, contemporary, historical, literary, fiction and non-fiction, written for different purposes and audiences — and the reading techniques that make an unglossed authentic text tractable without a dictionary.

5 sections·~35 min reading time·3 competencies·Level Standard 4 · Advanced 1

T·0222 / 12
Exam profile
AO2 · Understand and respond in writing to written Japanese from a variety of authentic sources and genresAO3 · Manipulate Japanese accurately when summarising and evaluating written sources in writingAO4 · Draw on knowledge of Japanese society and culture to read authentic texts in context
Operators:読みなさい答えなさい要約しなさい説明しなさい推測しなさい評価しなさい

basic level

At first-year standard the target is recovering main points and detail from a text of moderate length without panicking at unknown kanji, using the surrounding context and the structure of compounds. Reading widely and slowly matters more than reading fast; speed comes from familiarity with the themes, not from technique.

higher level

The full A-Level expects inference from complex material, awareness of a writer's purpose and audience, and the ability to summarise a position and weigh it against another. That means reading the sentence-final apparatus for stance, resolving 指示語 accurately across sentences, and distinguishing what a text asserts from what it reports.

Depth

Reading depth: In depth

Text

Text size: Standard

Contents · 5 sections▾
  1. Reading: Paper 1 Section B, and the reading half of Paper 3 Section B
    • 01Paper 1 Section B: four questions, forty-five minutes, answered in English◐
    • 02The text types the specification names, and how each one asks to be read◐
    • 03Reading an authentic text with no dictionary◐
    • 04The question types, and what each one is actually asking◐
    • 05The reading half of Paper 3 Section B, and weighing two sources●
§ 01

Paper 1 Section B: four questions, forty-five minutes, answered in English#

●●○StandardLPPearson Edexcel 9JA0 Paper 1 Section B — reading (20 marks, recommended 45 minutes)LPPearson Edexcel 9JA0 — language of assessment: Paper 1 Section B comprehension answered in EnglishLPDfE GCE AS and A level subject content for modern foreign languages

The forty-five-minute Section B routine

The Section B routineGraph, 設問 → 分類, 分類 → 走査, 走査 → 精読, 精読 → 英語設問分類走査精読英語
Fig. 1Five stations, in order, and the order is the whole point. Reading the questions before the text is counter-intuitive and is the single change that most improves timing, because it converts an open-ended reading task into a search with four targets. The emphasised edge runs into 精読, close reading, because that is the only station where the time should be spent: orientation and scanning are cheap, close reading is expensive, and a routine that reads everything closely spends its expensive minutes on paragraphs no question asks about. Node labels are held to two characters, since the figure engine measures label width in code points while a kanji draws at roughly twice that width.

Key points

Paper 1 Section B is 20 marks with a recommended 45 minutes and four questions, a mixture of multiple-choice and open-response. The texts are authentic or adapted from authentic sources; they contain both factual and abstract content; they include contemporary, historical, literary, fiction and non-fiction sources; and they are written for different purposes and audiences. Texts vary in length and the paper as a whole includes an extended passage. The section sits between the translation into English, which precedes it, and the research question, which follows it, and it is the middle of three sections in a two-hour-thirty-minute paper whose largest task is still to come — so overrunning here is expensive twice over.
The language rule is the one to fix first. The specification states that student responses to the translation in Paper 1 Section A and to the comprehension questions in Paper 1 Section B and Paper 3 Section A will be in English, and that all other student responses must be in Japanese. So Section B's questions are set in English and answered in English. This is a genuine difference from some sister qualifications and from Paper 3 Section B in this same qualification, and candidates do occasionally answer a Section B question in Japanese out of momentum. A Japanese answer to an English question does not earn extra credit; it risks not being credited at all.
Because the marks are distributed across four questions, triage matters more than raw reading speed. The routine that works has five stations. Read the question set FIRST, before the text, and classify each question by type — a fact question, a reason question, an inference question, a question about the writer's attitude. Then scan the text to locate the region each question belongs to, marking it. Then read those regions closely, in question order. Then write, in English, answering the question asked rather than reporting the sentence found. Then, if time permits, check the two things that most often go wrong: negation, and whether a claim belongs to the writer or to somebody the writer is quoting.
Two rules govern what a Section B answer should contain. First, answer the question asked and stop: a question asking for one reason wants one reason, and adding two more neither compensates for a wrong one nor earns extra credit, while it does consume minutes the research question needs. Second, do not lift undigested Japanese into an English answer. Copying 教育格差 into English prose leaves the examiner unable to tell whether you understood it; translate the term inside your own sentence, in the words the question needs. The converse failure — over-translating a term into a grand English abstraction the text does not support — is equally costly, and the middle path is to render the phrase plainly and move on.
Worked example

Triage across four questions, with the clock running

You have 45 minutes, 20 marks, four questions and two texts, one of them extended. Plan the section, then show how the plan changes when question 3 turns out to be hard.

  1. 01Minute 0 to 3: read all four questions and classify them

    Suppose question 1 is multiple-choice on the shorter text, question 2 asks for two reasons from the shorter text, question 3 asks what the writer implies about a policy in the extended text, and question 4 asks you to explain a phrase in context. Classification: 1 is recognition, 2 is retrieval, 3 is inference, 4 is close reading of a specific location. That tells you before you have read a word of Japanese that questions 1, 2 and 4 are locatable and question 3 is the one that will need thinking time.

  2. 02Minute 3 to 10: scan both texts and mark regions

    You are not reading for meaning yet. You are looking for the paragraph that mentions the policy in question 3, the phrase named in question 4, and the part of the shorter text that gives reasons for question 2. Japanese helps you here more than English would, because paragraph-initial connectives — しかし, 一方, つまり, たとえば — are visually distinctive and mark structural turns. Mark four regions, and note that they are unlikely to be evenly distributed: an extended passage often hides two answers in one paragraph.

  3. 03Minute 10 to 30: close-read the marked regions, in mark order

    Do the cheapest marks first: the multiple-choice, then the retrieval question, then the phrase in context. Each of those has a bounded search space and a bounded answer. That leaves question 3 with a known amount of remaining time rather than an unknown one, which is the whole purpose of ordering by cost. Write each answer as you finish it rather than collecting them for the end; an answer in your head at minute 44 is worth nothing.

  4. 04When question 3 turns out to be hard: budget, do not abandon

    Inference questions turn on a construction rather than on a word, and the constructions that carry implication are a small set: ~のではないか, ~とは限らない, ~わけではない, ~ざるを得ない, ~はずだ, ~に過ぎない. Find the one the paragraph uses and read outwards from it. If after five minutes you have not, write the most defensible partial answer you can from what you did understand and move on — an approximate English sentence may earn something and a blank cannot, which is the same logic as the translation section.

  5. 05Minute 40 to 45: the two checks that pay

    First, negation: re-read each of your answers against its source sentence and confirm that a negative in the Japanese is still a negative in your English, and that a double negative has not been flattened into a single one. Second, attribution: confirm that a claim you have attributed to the writer is not one the writer attributed to somebody else — ~という指摘がある and ~と言われている both mean the writer is reporting, not asserting. Those two errors account for a large share of avoidable Section B losses.

  6. 06And the discipline that makes the plan work

    Leave Section B at 45 minutes even if question 3 is unfinished. Paper 1 Section C is 40 marks — half the paper and 20 per cent of the whole A-Level — in a recommended 1 hour 25 minutes, and minutes borrowed from it to chase two marks here are the worst trade in the qualification. Write down the time you intend to leave each section before the examination begins, and treat it as a rule rather than as a guideline.

Result: The plan: three minutes to read and classify all four questions; seven to scan and mark the regions; twenty to close-read and answer in ascending order of cost, writing each answer as it is finished; five for a hard inference question capped by a hard stop; five for the negation and attribution checks. And the governing rule: leave at 45 minutes regardless, because Section C is worth twice this whole section and follows immediately.

Exam focus

  • Section B is answered in ENGLISH. Fix the rule before the examination, not during it (AO2).
  • Read the questions before the text and classify each by type (AO2).
  • Answer the question asked and stop. Extra content earns nothing and costs minutes (AO2).
  • Translate a term inside your own English sentence; never lift it and never inflate it (AO2).

Typical mistakes

  • Answering in Japanese out of momentum after reading a Japanese text.
  • Reading the whole text carefully before looking at the questions, and running out of time.
  • Writing three reasons when one was asked for, in the hope that one is right.
  • Copying a Japanese term into an English answer, which proves nothing about your understanding.

Active revision

日本語の記事を一つ選び、まず設問を四つ自分で作りなさい(事実を問うもの二つ、理由や筆者の考えを問うもの二つ)。次に、辞書を使わずに四十五分で、設問を先に読み、本文の該当箇所を探し、英語で答えなさい。最後に、各答えについて、本文のどの部分に基づいているかを日本語で示しなさい

Active recall

Recall the key points — then reveal.

Sources: Pearson Edexcel Level 3 Advanced GCE in Japanese (9JA0) specification (Pearson Edexcel)

§ 02

The text types the specification names, and how each one asks to be read#

●●○StandardLPPearson Edexcel 9JA0 Paper 1 Section B — texts contemporary, historical, literary, fiction and non-fiction, for different purposes and audiencesLPPearson Edexcel 9JA0 Paper 1 — texts vary in length and include an extended passageLPDfE GCE AS and A level subject content for modern foreign languages

Key points

The specification's list of text types is not decorative; each type has conventions that change how you should read it. A news report front-loads the facts and attributes them — ~によると, ~と発表した, ~とみられる — and its stance lives almost entirely in its verbs of attribution. An opinion column or 社説 states a position and argues it, so its structure is claim, ground, concession, restatement, and its connectives are load-bearing. An interview alternates question and answer, and the interviewer's questions frequently contain the summary of the previous answer, which makes them a free comprehension aid. A literary passage may leave subjects unstated across several sentences, which is the single hardest feature of Japanese narrative prose for a foreign reader.
Historical and older material can appear, and a small amount of preparation removes most of its difficulty. Pre-war and early post-war texts, and quotations within modern texts, may use 旧字体, the older kanji forms — 國 for 国, 學 for 学, 廣 for 広 — and may use 歴史的仮名遣い, historical kana spelling, where けふ is read きょう and さう is read そう. You are not expected to write any of this, and unlisted kanji will be glossed. What is worth knowing is that these forms exist and that an unfamiliar-looking character in an older text is very often a familiar one in an older shape, so the right instinct is to look for the modern equivalent rather than to assume an unknown word.
Purpose and audience change the vocabulary layer more than the grammar. Writing for a general newspaper audience uses 漢語 compounds heavily and keeps sentences long; writing for a younger audience or for online publication uses more katakana, more contraction and shorter sentences; official and administrative writing uses fixed formulas — ~に関する, ~に基づき, ~するものとする — that look forbidding and are in fact highly predictable once you have met a dozen of them. Identifying the register in the first two lines tells you which layer of vocabulary to expect and how much of an unfamiliar word you can safely guess.
Fiction deserves a specific warning because it is named in the specification and because it breaks the habits the other types build. Japanese narrative prose routinely omits subjects for long stretches, relies on the giving-and-receiving verbs to indicate who did what for whom, uses 心内語 — thought presented without a marker — and shifts between narrated time and remembered time without the tense signals English would use. The single most useful reading habit for such a passage is to track the subject explicitly in the margin, updating it whenever a giving-or-receiving verb, an honorific, or a ~てくれた tells you it has changed. That is the same skill the specification's error taxonomy demands in writing, read backwards.
Worked example

Reading a literary passage in which nobody is named

Read this and say, for each sentence, who is acting. 母が帰ってきたとき、台所には湯気が立っていた。何も言わずに鍋の蓋を開け、少しだけ味を見て、それから塩を足した。作ってくれたのだと分かって、急に恥ずかしくなった。台所を出るとき、後ろから、ありがとう、と聞こえた。

  1. 01Sentence 1: the only named participant, and a scene-setter

    母が帰ってきたとき、台所には湯気が立っていた. 母 is named and marked が because she is newly introduced. 帰ってきた uses くる to mark movement towards the speaker's location, which quietly tells you the narrator is at home. 湯気が立っていた describes the state of the kitchen with an intransitive verb and ている, so nobody is acting. Provisional margin note: narrator at home, mother arrives.

  2. 02Sentence 2: three verbs, no subject, and the default rule

    何も言わずに鍋の蓋を開け、少しだけ味を見て、それから塩を足した. No subject is stated. The default in Japanese narrative is that an unstated subject continues from the previous sentence's most salient participant — here 母, who has just arrived and acted. Nothing in the sentence contradicts that: the actions are ordinary kitchen actions and 何も言わずに, without saying anything, suits somebody who has just come in. Margin note: mother tastes and adds salt.

  3. 03Sentence 3: the giving verb reverses the reading

    作ってくれたのだと分かって、急に恥ずかしくなった. くれる means gives to me or to my side, so 作ってくれた means somebody made it FOR the narrator. That single verb tells you two things at once: the cooking was done by somebody other than the narrator, and the beneficiary is the narrator. And it forces a correction of sentence 2 — if the food had been made for the narrator by the mother, then the mother arriving and tasting is consistent, but 作ってくれた is past and refers to the making, so the pot was already cooking when she arrived. Someone else cooked.

  4. 04Resolve the contradiction by reading forward, not by guessing

    Sentence 4: 台所を出るとき、後ろから、ありがとう、と聞こえた. The narrator leaves and hears thanks from behind. Thanks is given by a beneficiary, and the person left in the kitchen is the mother — so the mother is thanking the narrator, and therefore the NARRATOR did the cooking, and 作ってくれた in sentence 3 must be the mother's perspective understood by the narrator, or a misreading. Re-read: the narrator realised that she had made it for him is one reading; the alternative is that 分かって belongs to the mother. Japanese permits both and the passage resolves it in the next paragraph.

  5. 05The lesson: mark the ambiguity rather than resolving it prematurely

    A foreign reader's instinct is to settle the subject and move on; the correct habit is to mark a subject as provisional and let subsequent evidence confirm or overturn it. In an examination this matters because an inference question will often turn precisely on the sentence where the subject changes. Writing a small question mark in the margin at sentence 3 costs nothing and preserves the option; committing silently to a reading costs the question.

  6. 06The three signals that carry subject change in Japanese narrative

    First, the giving-and-receiving verbs: くれる and もらう point towards the speaker's side, あげる away from it. Second, honorifics: an honorific verb elevates somebody other than the speaker, so it cannot describe the speaker's own action. Third, emotion and sensation predicates: 恥ずかしい, 嬉しい, 痛い in plain form describe the speaker's own state, and describing somebody else's requires ~がる or ~ようだ. Any of those three appearing in a sentence is a signal to check whose action you are reading.

Result: Sentence 1: the mother arrives; the kitchen state is described with no agent. Sentence 2: the unstated subject defaults to the mother. Sentence 3: くれる marks the narrator as beneficiary and 恥ずかしくなった marks the narrator as the one feeling, so the perspective is the narrator's throughout. Sentence 4: the thanks come from behind as the narrator leaves the kitchen, which points to the mother thanking the narrator — so the earlier attribution must be held provisionally. The method: mark subjects as provisional, and treat giving-and-receiving verbs, honorifics and emotion predicates as the three signals that a subject has changed.

Exam focus

  • Identify the text type in the first two lines and read according to its conventions (AO2).
  • Expect older kanji and kana forms in historical material; look for the modern equivalent (AO2).
  • In an interview text, use the interviewer's questions as free summaries of the previous answer (AO2).
  • In fiction, track the unstated subject in the margin, updating at every giving-or-receiving verb (AO2).

Typical mistakes

  • Reading a news report for opinion. Its stance is in its attributions, not in adjectives.
  • Treating an older kanji form as an unknown character rather than as a familiar one in an older shape.
  • Losing the subject in a literary passage and attributing an action to the wrong character.
  • Assuming administrative formulas are hard. They are fixed, few, and worth memorising as a set.

Active revision

四種類の日本語の文章(新聞記事、社説、インタビュー、小説の一節)をそれぞれ一つずつ集めなさい。各文章について、(一)筆者の立場がどこに現れているか、(二)主語が省略されている箇所がいくつあるか、(三)その主語をどう特定したか、を日本語で説明しなさい

Active recall

Recall the key points — then reveal.

Sources: Pearson Edexcel Level 3 Advanced GCE in Japanese (9JA0) specification (Pearson Edexcel)

§ 03

Reading an authentic text with no dictionary#

●●○StandardLPPearson Edexcel 9JA0 — the use of dictionaries is not permitted in any paperLPPearson Edexcel 9JA0 Appendix 3 — kanji list; off-list kanji glossed in the papersLPDfE GCE AS and A level subject content for modern foreign languages

Key points

No dictionary is permitted in any paper of this qualification, so every text is a closed system and the only resources are the ones already in your head plus the text's own internal evidence. That sounds harsher than it is, because two facts work in your favour. First, Appendix 3 tells you which kanji you are expected to recognise, and any kanji used in a question paper that is not on the list will be glossed. Second, Japanese expository writing glosses its own terminology far more readily than English does: ~という, ~とは, つまり, 言い換えれば and すなわち all introduce a definition, and a term of art in a well-written Japanese article is very often defined within a sentence of its first appearance.
For an unfamiliar compound, the structure of the compound itself is the first resource. Most 漢語 compounds are two characters whose relationship is one of a small number of types: modifier plus head (高齢化 — old age plus becoming), verb plus object (読書 — reading plus books), coordination of near-synonyms (困難 — difficulty plus difficulty), or coordination of opposites (increase and decrease, 増減). Knowing the individual characters therefore usually yields a workable approximation. 過疎[かそ] combines excessive with sparse and means depopulation; 空洞化 combines empty and cavity with becoming and means hollowing-out. You do not need the dictionary meaning; you need enough to answer the question.
Katakana is a separate and easier problem, because it is phonetic. A katakana word you do not recognise is very often an English word in Japanese phonology, and sounding it out recovers it: インフラ, サステナブル, ガイドライン, オーバーツーリズム. Two habits help. Remember that long vowels are marked with ー and that a final consonant usually acquires a vowel, so ストレス is stress and ビジネス is business. And remember that some katakana words are not English at all — アルバイト is from German, パン from Portuguese — and that some are Japanese abbreviations of English phrases whose full form was never used, such as コンビニ and リモコン.
The last resource is the sentence's own grammar, and it is the most reliable of the three. Even with an unknown noun, the particles tell you its role, the verb tells you what happened to it, and the sentence-final apparatus tells you the writer's stance towards the whole. A sentence such as 制度の硬直化が問題視されている is fully readable even if 硬直化 is unknown: が marks it as the subject, 問題視される means to be regarded as a problem, ている makes it a standing situation, and the passive leaves the regarders unnamed. You now know that some property of the system is widely regarded as a problem, which will answer most questions that could be asked about the sentence.
Worked example

One reading decision, walked through: 指示語 and what それ points at

Resolve every 指示語 in this passage, and say how you know. 地方の自治体では、若い世代の流出が続いている。これに対して、住宅の補助や仕事の紹介などの対策が取られてきた。しかし、そうした取り組みが効果を上げた例は多くない。その理由として、雇用そのものが増えていないことが挙げられる。これは、対策の設計の問題というより、地域経済の構造の問題である。

  1. 01これに対して: これ points at a situation, not a noun

    これ here refers to the whole preceding clause — the continuing outflow of young people — not to any single noun in it. This is the commonest use of これ in expository prose and it catches readers who look for a noun antecedent. The test is grammatical: に対して takes a thing responded to, and the response described is a set of policies, which respond to a situation rather than to a generation. Read これ as the state of affairs just described unless a specific noun is clearly indicated.

  2. 02そうした取り組み: そうした restricts, and the restriction matters

    そうした取り組み means initiatives of that kind — specifically the housing subsidies and job introductions just listed, not initiatives in general. That restriction is doing argumentative work: the sentence claims that THOSE measures have rarely succeeded, which is a narrower and more defensible claim than that nothing works. A reader who takes そうした as merely such would flatten the claim, and a question asking what exactly the writer says has failed would be answered wrongly.

  3. 03その理由: その points back one sentence, to a fact just asserted

    その理由 is the reason for that, where that is the previous sentence's claim — that few such initiatives have had much effect. Note that その and これ are doing different jobs in this passage: これ pointed back to a situation across a sentence boundary at the start, and その points back to a proposition the writer has just advanced. The rough working rule is that これ points to what is close or newly introduced and その to what has just been established, and in expository writing その is far commoner.

  4. 04これは: the second これ, and why it is not the same as the first

    これは、対策の設計の問題というより、地域経済の構造の問題である. This これ refers to the fact just given as the reason — that employment itself is not growing. The writer is classifying that fact: it is a structural problem rather than a design problem. ~というより means rather than, and it is a comparative rejection, not a denial: the writer is not saying the design is perfect, only that the design is not the main issue. Reading というより as a flat negative is a standard error.

  5. 05Build the chain, and check it holds

    Outflow continues → these policies were adopted in response → policies of that kind have rarely succeeded → the reason is that employment is not growing → and that is structural rather than a design fault. Every link is carried by a 指示語, and the paragraph would be incomprehensible without resolving them. Reading the chain out as a numbered list, as here, is a two-minute exercise that turns a dense paragraph into an argument you could summarise in one sentence.

  6. 06Why this matters for the marks

    Section B questions ask what the writer thinks, what a phrase refers to, and why the writer holds a view — all three of which are answered by the 指示語 chain rather than by any single sentence. And Paper 3 Section B's 5(b) asks you to summarise a written source's point of view, which is precisely the chain, not the sentences. So resolving 指示語 explicitly is not a comprehension exercise for beginners; it is the main technical skill that reading in this qualification tests.

Result: これに対して → the continuing outflow of young people, a situation and not a noun. そうした取り組み → specifically the housing subsidies and job introductions just listed, a restricted set. その理由 → the reason for the claim just made, that few such measures have succeeded. これは → the fact that employment itself is not growing. And というより → a comparative rejection, not a denial. The chain is the argument, and summarising the passage means summarising the chain.

Exam focus

  • Look for the text's own gloss: ~とは, ~という, つまり, 言い換えれば (AO2).
  • Decompose an unknown 漢語 compound into its characters and their relationship (AO2).
  • Sound out unfamiliar katakana; most of it is borrowed and phonetic (AO2).
  • Read the grammar around an unknown word: particles, verb and sentence-final apparatus (AO2).

Typical mistakes

  • Stopping at an unknown word instead of reading the sentence's grammar around it.
  • Assuming every katakana word is English. アルバイト, パン and カルテ are not.
  • Guessing a compound from one of its characters instead of from the relationship between both.
  • Ignoring the gloss the writer supplied one line later, because you stopped reading at the term.

Active revision

日本語の記事を一つ選び、知らない語に丸をつけなさい。各語について、(一)漢語なら字の組み合わせから、(二)カタカナなら音から、(三)どちらでもなければ前後の文法から、意味を推測し、その根拠を日本語で書きなさい。辞書は最後に一度だけ使い、推測がどれだけ当たったかを確認しなさい

Active recall

Recall the key points — then reveal.

Sources: Pearson Edexcel Level 3 Advanced GCE in Japanese (9JA0) specification (Pearson Edexcel)

§ 04

The question types, and what each one is actually asking#

●●○StandardLPPearson Edexcel 9JA0 Paper 1 Section B — four questions, multiple-choice and open-responseLPPearson Edexcel 9JA0 Paper 1 — infer meaning from complex written material; assimilate and use information from written sourcesLPDfE GCE AS and A level subject content for modern foreign languages

Key points

Multiple-choice questions in a foreign-language reading test are not the soft option they look like, because the distractors are built from the text. A typical set will contain one option that is true of the text but does not answer the question, one that contains a real phrase from the text used with the wrong scope, one that reverses a negative, and one that is correct. Recognising those four shapes is most of the skill. The working method is to answer the question from the text FIRST, in your own words, and only then to look at the options and find the one that matches — reading the options first invites you to recognise familiar Japanese rather than to check meaning.
Open-response questions divide into retrieval and inference, and the verb of the question tells you which. State, give, name and according to the text signal retrieval: the answer is in the text and your job is to find it and render it. Suggest, imply, why does the writer and what does the writer think signal inference: the answer is carried by a construction rather than stated, and you will find it in the sentence-final apparatus, in a concessive, or in a 指示語 chain. Misclassifying an inference question as retrieval produces an answer that quotes a sentence and does not answer; misclassifying a retrieval question as inference produces an answer that speculates when the text was explicit.
The instruction in your own words appears in reading tests and means something specific when the source language differs from the answer language. Because Section B is answered in English and the text is in Japanese, every answer is already in your own words in the trivial sense. What the instruction is guarding against is the two failures named earlier: lifting a Japanese term untranslated, and producing a word-for-word rendering that reads as a translation exercise rather than as an answer. The safe practice is to state the answer as a claim in ordinary English and, where a specific term matters, to gloss it in passing — the private after-school schools known as juku — rather than to import it raw.
Two structural habits are worth building into every answer. First, answer at the grain the question asks for: a question worth one mark wants one point, and a question worth three wants three, and matching the number of points to the number of marks is a free and reliable check. Second, when a question asks why, make sure your answer contains a because or an equivalent — an astonishing number of why answers are descriptions that never state a reason. Reading your own answer back and asking whether it answers the interrogative word takes five seconds and catches a whole class of error.
Worked example

A multiple-choice question and an inference question on the same paragraph

The paragraph is: 若者の車離れがよく話題になるが、調査によれば、車を持ちたくないというより、持てない事情がある場合が多いという。都市部では駐車場の費用が高く、公共交通も発達している。ただし、地方では事情が異なり、車がなければ生活が難しい地域も少なくない。 Question 1, multiple-choice: According to the text, why do many young people not own cars? (a) They dislike driving. (b) In cities, costs and good public transport make it unnecessary or unaffordable. (c) Cars are unavailable in rural areas. (d) Surveys show attitudes have not changed. Question 2, open-response: What does the writer imply about the phrase 若者の車離れ?

  1. 01Question 1: answer from the text before looking at the options

    The text's own answer is in 持ちたくないというより、持てない事情がある — not that they do not want to, but that circumstances prevent it — followed by the two circumstances: parking costs are high in cities and public transport is developed. So the answer, in your own words before reading any option, is: because in cities it is expensive and public transport makes a car unnecessary. Now look at the options and (b) matches. Total time: under a minute, and no distractor had a chance to work on you.

  2. 02See what each distractor was built from

    (a) dislike driving is the reversed version of 持ちたくないというより — the text explicitly rejects it, which is the reverse-the-negative distractor. (c) uses 地方 and 車 from the last sentence but with the scope inverted: the text says cars are NECESSARY in rural areas. (d) uses 調査によれば from the text, attached to a claim the text never makes. Each distractor is made of real material used wrongly, which is why answering from the text first is the defence.

  3. 03Question 2: classify it — implies means inference

    The verb is implies, so the answer is not stated. Look for the construction: が in the first sentence is the contrastive but, and it sets 若者の車離れがよく話題になる against what the survey found. That structure — this is often talked about, BUT the evidence says something else — implies that the writer regards the phrase as a common framing that the evidence does not support. That is an implication carried by が and by というより, not by any noun.

  4. 04Check the second construction before answering

    持ちたくないというより、持てない事情がある. 車離れ literally means a moving away from cars and connotes a change of taste — a generation that has turned away. というより rejects precisely that connotation and substitutes constraint. So the implication is specifically about the word: it attributes a preference where the writer says circumstances are at work. Naming which part of the phrase the writer is contesting is what turns a general answer into a precise one.

  5. 05Write the inference answer in English, and keep it to the implication

    That the phrase is misleading: it suggests young people have chosen to turn away from cars, whereas the writer says the evidence points to circumstances that prevent ownership, particularly cost and available alternatives in cities. That answers implies, names what is misleading, and cites the ground. It does not say the writer thinks young people love cars — the text does not say that — and it does not translate the paragraph.

  6. 06The check that catches the standard error here

    Read your answer and ask: is it about the PHRASE, or about car ownership? The question asked what the writer implies about 若者の車離れ, which is a phrase, so an answer about how many young people own cars has answered a different question. Questions about a word or a phrase are common in this section precisely because they test whether you can separate what a text says from how it frames what it says, and that separation is what the AO2 grid means by inferring meaning from complex written material.

Result: Question 1: option (b), reached by answering from the text before reading the options, with the three distractors identifiable as a reversed negative, a scope inversion and a false attachment to 調査によれば. Question 2: that the phrase is misleading — it implies a change of taste, whereas the writer says the evidence points to circumstances, especially cost and public transport in cities, that prevent ownership. The implication is carried by the contrastive が and by というより, and the answer must be about the phrase rather than about car ownership.

Exam focus

  • Answer a multiple-choice question from the text first, then match to an option (AO2).
  • Classify each open-response question by its verb: retrieval or inference (AO2).
  • Gloss a Japanese term inside your English sentence; never import it raw (AO2).
  • Match the number of points in your answer to the number of marks, and answer the interrogative word (AO2).

Typical mistakes

  • Reading the multiple-choice options first and picking the one containing familiar Japanese.
  • Answering an inference question by quoting the sentence that contains the topic.
  • Speculating on a retrieval question when the text stated the answer plainly.
  • Writing a description in answer to a why question, with no reason anywhere in it.

Active revision

日本語の記事を一つ選び、同じ本文について四択問題を一つ作りなさい。誤答の三つは、(一)本文に書いてあるが設問に答えていないもの、(二)本文の語句を範囲を変えて使ったもの、(三)否定を逆にしたもの、とすること。そのうえで、自分で解き、なぜその誤答が魅力的かを日本語で説明しなさい

Active recall

Recall the key points — then reveal.

Sources: Pearson Edexcel Level 3 Advanced GCE in Japanese (9JA0) specification (Pearson Edexcel)

§ 05

The reading half of Paper 3 Section B, and weighing two sources#

●●●AdvancedLPPearson Edexcel 9JA0 Paper 3 Section B — summarise the different points of view in each source, then evaluateLPPearson Edexcel 9JA0 Paper 3 Section B — set in Japanese and answered in JapaneseLPDfE GCE AS and A level subject content for modern foreign languages

Key points

In Paper 3 Section B the reading task changes shape twice over. First, the output is Japanese rather than English: 5(b) asks you to summarise the written source's points of view in Japanese, and 5(c) asks for an evaluation in Japanese of approximately 400 to 500 characters. Second, the reading is comparative by design, because 5(c) asks you to evaluate the viewpoints in BOTH sources and draw conclusions, and the AO1 grid out of 6 and the AO2 grid out of 6 are printed side by side precisely because the two sources have to be weighed against each other. So you are not reading for content alone; you are reading for the position a source takes and for the grounds it takes it on.
That changes what you underline. In Paper 1 Section B you mark the regions where answers live. Here you mark the source's claim, its grounds, its concessions and its hedges, because those four things are what a summary of a point of view consists of and what an evaluation has to weigh. A text that concedes something is easier to summarise fairly and harder to dismiss; a text that hedges — ~可能性がある, ~かもしれない, ~とは限らない — is claiming less than a careless reader will report it as claiming, and reporting it as claiming more is a misreading that both the AO2 grid and the evaluation will punish.
The summary at 5(b) is points-based, like 5(a), so the same discipline applies: short attributed sentences, every distinct viewpoint recovered, no evaluation. The reporting verbs available for a written source differ slightly from those for a speaker and are worth having: ~と述べている for states, ~と主張している for argues, ~を指摘している for points out, ~を挙げている for cites, ~と結論づけている for concludes. Using several rather than repeating one also feeds the AO3 grid, which in this section is a combined accuracy-and-range grid out of 10 covering the whole of your Section B writing.
For 5(c), the highest-value move is the one taught in the artistic-culture and social-issues topics: locate the disagreement precisely before evaluating it. Do the two sources disagree about facts, about causes, about what should be done, or about who should bear a cost? Very often they agree on more than they appear to, and naming the shared ground first — 二つの資料は、事実については対立していない — proves that you understood both, which is exactly what the AO1 and AO2 grids measure. Then name the criterion on which the question turns, apply it, and conclude at the strength your reasons support.
Worked example

5(b): summarising a written source in Japanese, with its hedges preserved

The written source argues that raising the retirement age would help with labour shortage; concedes that older workers' health varies considerably; cites a survey showing many people want to keep working; and concludes, with a hedge, that this could be part of a solution but is not sufficient on its own. Write 5(b).

  1. 01Identify the four elements before writing anything

    Claim: raising the retirement age would help. Grounds: a survey showing many want to keep working. Concession: health varies. Hedge: could be part of a solution, not sufficient alone. Those four map onto four sentences, and writing them in that order produces a summary whose shape mirrors the source's argument — which is what summarising a point of view means, as opposed to listing what a text mentions.

  2. 02Sentence one: the claim, with the right reporting verb

    この文章は、定年の延長が人手不足の緩和につながると主張している. 主張する is argues, stronger than 述べる and correct for a text advancing a case. 定年の延長 is the raising of the retirement age; 緩和[かんわ] is easing; ~につながる is leads to, a very useful verb for causal claims because it is weaker than 引き起こす and therefore usually more accurate. Note that この文章は identifies the source, which matters when 5(c) will compare it to another.

  3. 03Sentence two: the grounds, with a different verb

    その根拠として、働き続けたいと考える人が多いという調査結果を挙げている. ~を根拠として挙げる is cites as grounds, and it is a different verb from the first sentence. 働き続けたいと考える人 is a relative clause using たい plus と考える. 調査結果 is survey findings. The whole sentence attributes the survey to the text rather than presenting it as fact, which is correct: you have not seen the survey and neither has the examiner.

  4. 04Sentence three: the concession, marked as a concession

    一方で、高齢者の健康状態には個人差が大きいことも認めている. 一方で marks the turn; 個人差[こじんさ] is individual variation, a precise noun; ことも認めている uses 認める, acknowledges, which is the verb reserved for a point conceded against the writer's own case. Including the concession is not padding — a summary that omits it reports the source as more one-sided than it is, and the AO2 grid is about understanding the source.

  5. 05Sentence four: the hedge, at its own strength

    結論として、定年の延長は対策の一つにはなりうるが、それだけでは十分ではないとしている. ~うる after a ます-stem means can be, and なりうる is could become — the precise rendering of could be part of. だけでは十分ではない is not sufficient on its own. ~としている is a neutral attribution verb for a text's stated position. Reporting a hedged conclusion as a firm one would misrepresent the source, and a 5(c) built on that misreading would then attack a position nobody holds.

  6. 06Check: four elements, four verbs, no evaluation

    主張している, 挙げている, 認めている, としている. Four reporting verbs, none repeated, feeding the range half of the combined AO3 grid at no cost. And nothing in the summary says whether the argument is any good, which is correct: evaluation belongs at 5(c), where its marks are, and a candidate who evaluates here has spent characters where no marks exist. Read it back once for the hedge — is the conclusion still tentative in your summary? It is.

Result: この文章は、定年の延長が人手不足の緩和につながると主張している。その根拠として、働き続けたいと考える人が多いという調査結果を挙げている。一方で、高齢者の健康状態には個人差が大きいことも認めている。結論として、定年の延長は対策の一つにはなりうるが、それだけでは十分ではないとしている。 Claim, grounds, concession and hedge — the four elements of a point of view, in the source's own order, with four different reporting verbs and no evaluation.

Exam focus

  • Everything in Paper 3 Section B is answered in Japanese, including the summaries (AO2).
  • Underline claim, grounds, concession and hedge — those four things ARE the point of view (AO2).
  • Vary the reporting verbs: 述べている, 主張している, 指摘している, 挙げている, 結論づけている (AO3).
  • At 5(c), name the shared ground before the disagreement; it proves you read both sources (AO1, AO2).

Typical mistakes

  • Summarising a hedged claim as though it were asserted. ~可能性がある is not the same as である.
  • Repeating と述べている five times when four other reporting verbs are available and markable.
  • Evaluating at 5(b). The summary carries no evaluation marks and the characters are needed at 5(c).
  • Jumping to which source you agree with without naming the criterion that decides it.

Active revision

同じ話題について立場の異なる日本語の文章を二つ用意しなさい。それぞれについて、(一)主張、(二)根拠、(三)譲歩、(四)ぼかし表現、の四点に線を引きなさい。次に、各文章を三文以内の日本語で要約し、伝達表現は毎回変えること。最後に、二つの資料が一致している点を三つ挙げてから、対立の軸を一文で書きなさい

Active recall

Recall the key points — then reveal.

Sources: Pearson Edexcel Level 3 Advanced GCE in Japanese (9JA0) specification (Pearson Edexcel)

Contents

Section -- / 05

    • 01Paper 1 Section B: four questions, forty-five minutes, answered in English◐
    • 02The text types the specification names, and how each one asks to be read◐
    • 03Reading an authentic text with no dictionary◐
    • 04The question types, and what each one is actually asking◐
    • 05The reading half of Paper 3 Section B, and weighing two sources●

0/5 Read

From notes into training

Reading: Paper 1 Section B, and the reading half of Paper 3 Section B

Reinforce this topic with matching tasks from the question bank.

~35
min
3
Competencies
Practise

References & sources

Sources

Pearson Edexcel

  • Pearson Edexcel Level 3 Advanced GCE in Japanese (9JA0) specification

Previous topic

Grammar and translation (Paper 1 Section A, Paper 2 Section A, Appendices 2 and 3)

Next topic

Listening: Paper 3, and the whole of the AO1 weighting

EuraStudy·Notes T·02·MMXXVI

Carry on to the next topic — your learning path is kept.