মৌখিক পরীক্ষা নেই: spoken Bengali in a written qualification

Using only the specification's own figures, work out how the 400 scaled marks are distributed across the papers and across the tasks inside them, and decide what a revision week should look like.

1. 01Scale the three papers

   Paper 1 has a maximum raw mark of 85 and a scaling factor of two, giving 170 scaled. Paper 2 has 80 raw and a factor of one, giving 80. Paper 3 has 75 raw and a factor of two, giving 150. Total 400.

2. 02Scale the tasks inside Paper 1

   Reading 35, translation into English 10, research essay 40 - doubled, that is 70, 20 and 80 scaled.

3. 03Scale Paper 2 and Paper 3

   Paper 2 is two essays of 40 each, unscaled: 40 and 40. Paper 3 is listening 35, translation into Bengali 10 and multi-skill task 30 - doubled, that is 70, 20 and 60.

4. 04Rank the tasks by scaled marks

   Research essay 80; listening 70; reading 70; multi-skill task 60; each works essay 40; each translation 20. Writing tasks together carry 220 of the 400 scaled marks; comprehension carries 140; the two translations carry 40.

5. 05Draw the conclusion for a revision week

   More than half the qualification is essay writing in Bengali, and 140 scaled marks depend on comprehension of speech and text. Nothing at all depends on speaking. So a week that contains four hours of writing practice, three of listening, three of reading and thirty minutes of reading aloud is proportionate; a week built around conversation practice is not, however enjoyable.

6. 06Note the one exception

   That is the exam arithmetic, and it is not the only reason to study a language. If you want to speak Bengali with your family, spend the time - just do not call it revision, and do not let it displace the essay practice that 220 scaled marks depend on.

Result: Scaled marks: Paper 1 170, Paper 2 80, Paper 3 150, total 400. By task: research essay 80, reading 70, listening 70, multi-skill task 60, each works essay 40, each translation 20. Nothing is awarded for speech. The arithmetic settles the revision plan, and it settles it against the intuition most language students bring with them.

Build a five-minute daily drill that trains the contrasts you actually need, using vocabulary from the specification's themes rather than from a phonetics textbook.

1. 01Take the aspirated pairs from real words

   কাজ, work, against খবর, news; গান, song, against ঘর, house; তারিখ, date, against থানা, police station; পরিবার, family, against ফল, result or fruit. Say each pair aloud five times, exaggerating the breath on the aspirated member.

2. 02Take the dental-retroflex contrast from real words

   দিন, day, against ডাক, call; তালিকা, list, against টাকা, money; নদী, river, against ণ, which appears only inside words such as কারণ. Feel the tongue move from the teeth to the roof of the mouth.

3. 03Take the flap from theme vocabulary

   বড়, big; পড়া, to read or study; ঝড়, storm; সড়ক, road - a word you will meet in any passage about the road-safety movement. Say each with a flap, not a stop.

4. 04Take the doubling conjuncts from examination vocabulary

   পরীক্ষা, examination; বিশ্ববিদ্যালয়, university; স্বাধীনতা, independence; সত্য, truth; বিদ্যালয়, school; জ্ঞান, knowledge. These six words appear constantly in this specification's material and all six contain a conjunct that is not sounded as written.

5. 05Add nasalisation and the velar

   চাঁদ, moon; দাঁড়ি, the full stop; বাংলা; আন্দোলন, movement. Two nasal systems, four words, thirty seconds.

6. 06Close by listening

   Finish every drill by hearing the same words in a recording, so that the connection between what you just produced and what the exam will play is made immediately rather than left to chance.

Result: A five-minute drill of about twenty-five words, every one of them drawn from vocabulary the specification's themes actually use. The drill trains four contrasts, costs nothing, is never assessed, and pays into the 70 scaled marks of the listening section.

You need a paragraph on why young people migrate. Compose it aloud first, then write it, and record what the aloud pass caught.

1. 01Say it once, at speed, without stopping

   গ্রামে কাজ নেই, তাই অনেক তরুণ শহরে যায়। তিন চার বছর পরে কেউ কেউ ফিরে আসে, কিন্তু অধিকাংশই থেকে যায়। Notice what happened: you produced two short sentences with a causal connective, which is a better shape than most silently drafted paragraphs achieve.

2. 02Hear the first error

   তিন চার বছর has no classifier. Said aloud, it sounds incomplete to a Bengali ear in a way it does not look incomplete on the page. Corrected: তিন-চার বছর is in fact acceptable for a duration, but তিন চার তরুণ would not be - and hearing the gap is what alerts you to check.

3. 03Hear the second error

   অধিকাংশই থেকে যায় - the emphatic -ই on অধিকাংশ is right, but said aloud you may notice that থেকে যায় is a compound verb meaning to stay on, which is exactly what you want; had you written রয়ে যায় or just থাকে the meaning would have shifted. The aloud pass is where verb choice gets tested against the ear.

4. 04Extend aloud before writing

   কারণ শহরে যত কম বেতনই হোক, নিয়মিত আয় থাকে। Because however low the pay in the city, there is a regular income. যত … হোক is a concessive construction that is much easier to produce aloud than to assemble on paper.

5. 05Now write it, and check the three things speech cannot check

   Spelling, the নুক্তা and the চন্দ্রবিন্দু. Speech is silent about orthography, so the written pass must check exactly what the aloud pass could not: বেতন, আয়, অধিকাংশ, and any নুক্তা in the paragraph.

6. 06Note the division of labour

   The aloud pass catches structure, verb choice, honorific level and rhythm. The written pass catches spelling, diacritics and punctuation. Neither substitutes for the other, and doing them in that order is faster than doing either twice.

Result: গ্রামে কাজ নেই, তাই অনেক তরুণ শহরে যায়। তিন-চার বছর পরে কেউ কেউ ফিরে আসে, কিন্তু অধিকাংশই থেকে যায়, কারণ শহরে যত কম বেতনই হোক, নিয়মিত আয় থাকে। Composed aloud in about forty seconds, written in two minutes, and containing one concessive construction that silent drafting rarely produces.

You believe that remittances have improved life in rural Bangladesh. Argue it, take the objections, and write what survives.

1. 01State the claim as you would in an essay

   প্রবাসী আয় গ্রামীণ বাংলাদেশে জীবনযাত্রার মান উন্নত করেছে। Remittances have improved living standards in rural Bangladesh. Clear, arguable, and immediately vulnerable - which is what you want.

2. 02Take the first objection: what about the cost of going?

   কিন্তু যাওয়ার খরচ কে দিয়েছে? Many households borrowed or sold land to pay recruitment costs, so the first years of remittance repay the journey. The claim survives but must be narrowed: the improvement is real but delayed and conditional on the placement succeeding.

3. 03Take the second objection: improved compared with what?

   উন্নত - compared to what counterfactual? If incomes rose everywhere over the same period, the remittance effect is smaller than it looks. The honest answer is that separating the two requires evidence you do not have, and saying so is better than pretending.

4. 04Take the third objection: improved for whom?

   The household with a migrant is not the village. If remittance income raises local prices, households without a migrant may be worse off. This is a real argument made by researchers and it is the objection that most improves the paragraph.

5. 05Write the claim that survives

   প্রবাসী আয় যেসব পরিবারে আসে, সেখানে ঘরবাড়ি, শিক্ষা ও চিকিৎসায় স্পষ্ট উন্নতি দেখা গেছে; তবে এই উন্নতি অভিবাসনের ঋণ শোধ হওয়ার পরেই শুরু হয়, এবং যেসব পরিবারে কেউ বিদেশে যায়নি, তাদের অবস্থার সঙ্গে তুলনা করা কঠিন। কিছু গবেষক মনে করেন এতে গ্রামের ভেতরেই নতুন বৈষম্য তৈরি হয়েছে।

6. 06Notice what changed

   The original claim was about a country; the surviving claim is about households, with a time condition, an acknowledged comparison problem and an attributed counter-position. It is a much smaller claim and worth many more marks, because the AO4 grid rewards evidence that justifies a conclusion, and this conclusion is one the evidence actually reaches.

Result: প্রবাসী আয় যেসব পরিবারে আসে, সেখানে ঘরবাড়ি, শিক্ষা ও চিকিৎসায় স্পষ্ট উন্নতি দেখা গেছে; তবে এই উন্নতি অভিবাসনের ঋণ শোধ হওয়ার পরেই শুরু হয়, এবং যেসব পরিবারে কেউ বিদেশে যায়নি, তাদের অবস্থার সঙ্গে তুলনা করা কঠিন। কিছু গবেষক মনে করেন এতে গ্রামের ভেতরেই নতুন বৈষম্য তৈরি হয়েছে। Three objections, one narrowed claim, and a paragraph that could not have been written by someone who had not been argued with.

You speak Bengali at home and have never written it formally. Work out, in one afternoon, exactly what you need to learn.

1. 01Test one: dictation

   Have somebody read you a hundred words of newspaper Bengali. Write it. Correct it. Count the errors and classify them: sibilants, ণ versus ন, conjuncts, নুক্তা, চন্দ্রবিন্দু, other. This single test will usually reveal that four or five specific patterns account for most of your spelling errors, and those patterns are what you practise.

2. 02Test two: register

   Write one paragraph on a thematic question, then underline every word you would use in conversation at home. If most of the paragraph is underlined, the register is spoken and needs raising. Replace ten of the underlined items with formal equivalents - বলা with উল্লেখ করা, দেখা with পর্যবেক্ষণ করা, ভাবা with বিবেচনা করা - and read the difference.

3. 03Test three: dialect

   Ask somebody who writes standard Bengali to read the same paragraph and mark only what is dialectal rather than wrong. Sylheti and Chittagonian speakers in particular will find specific verb forms and vocabulary items flagged. Write the list; it will be short and it will not grow.

4. 04Test four: metalanguage

   Try to write, in Bengali, one sentence about how a film director uses the camera and one about how a narrator's voice affects a story. If you cannot, you have found the Paper 2 gap - and it is a vocabulary gap, not an ideas gap, which means it is closable in a fortnight.

5. 05Test five: the other convention

   Read a page from the country you did not grow up with. Mark every word you had to think about. If জল, কাকা, মাসি or নুন slowed you down, or পানি, চাচা, খালা and লবণ did, that is a reading list rather than a problem.

6. 06Write the plan

   Five tests, five short lists, and a term's worth of specific work: dictation twice a week for the spelling patterns; ten formal replacements a week for register; the dialect list on a card; the analytical vocabulary as a set; and one text a week in the other convention.

Result: An afternoon of testing produces five short, specific lists and replaces the vague anxiety of a fluent speaker who keeps losing marks with a term of targeted work. The gaps are narrow, they are known, and none of them is about speaking - which is fortunate, because nothing spoken is assessed in this qualification.