High Marks in English, No English: The Translation-and-Dictation Routine
In a Class V classroom of 36 children in Agartala, English is the third language for most. The teacher's daily routine is fixed: he reads the English paragraph aloud, translates each sentence into Bengali, then dictates the answers to the textbook questions, which the children memorise for the test. In the last monthly test, 28 of 36 scored above 60%, almost all by reproducing those answers. When the teacher greeted the class in English and asked, 'What is your name?', three children answered, and two of those three were the children of English-medium tutors. A quick check showed that 24 children read 'mat' and 'mate' identically and 30 could not write a single sentence about their own family without copying. Sanjoy, whose home has a cable connection, reads English storybooks, finishes the copied work in five minutes and then disturbs the class. The head teacher treats the high test scores as proof that the method works.
1. Is the head teacher right that the test scores prove the method works? What exactly does the evidence show?
Read the worked answer
No. The test is measuring reproduction, not English.
- 28/36 pass, but the answers were dictated and memorised. A test that rewards recall of a dictated string tells you the string was stored, nothing more. The transfer test is different: only 3/36 answered a routine greeting, and 30/36 could not write one self-generated sentence. Those two results, taken together, are the real finding.
- Decoding is absent. 24 children read 'mat' and 'mate' identically. English vowel graphemes are not being processed; the children are matching whole word shapes to memorised sounds. This caps all future progress - they cannot read an unseen word.
- Translation of every line removes the need to process English. The child waits for the Bengali, so the English never has to be understood.
- Sanjoy's disruption is information. He is the only fluent reader, and his task is finished; the problem is a mismatch between task difficulty and his level, not indiscipline.
What would count as evidence of learning: a child reads an unseen three-word sentence aloud, answers a question she has not seen before, and writes one true sentence about herself with support. Those are the tests to run.
2. Restructure a 40-minute English period so that both beginners and Sanjoy are working. Give the time blocks and the reasoning behind each.
Read the worked answer
Block 1 – 7 min: Sounds and sight words (whole class, mixed pairs). Explicit phonics of one grapheme pair (a_e /a/ as in 'mate', 'cake'), then five sight words on cards. Beginners match word to picture; stronger children sort by sound. Pairs are mixed so Sanjoy or another fluent child is a reading partner - this gives him a role instead of an audience.
Block 2 – 10 min: Oral English, no writing. TPR and a substitution drill on one frame: 'My name is ___. I live in ___.' Children say it to three partners. Speaking before writing removes the transcription barrier.
Block 3 – 12 min: Reading. Beginners: a four-line text with pictures, three literal questions answered orally. Sanjoy's group: a short unseen passage at his level with two inferential questions ('Why do you think…?'). Same skill, different text - that is differentiation, not extra work.
Block 4 – 8 min: Writing. One sentence from the frame, with a word bank on the board. Beginners copy the frame and change one word; Sanjoy's group writes three sentences without the bank.
Block 5 – 3 min: Close. Two children read their sentence aloud. No dictation of test answers at any point.
3. How should the teacher handle spoken errors during Block 2, and why?
Read the worked answer
Correct selectively, and mostly by recasting.
- During a fluency task (Block 2), the goal is that every child speaks. Stopping a child mid-sentence to correct 'I is Sanjoy' makes the speaking task feel like a trap; the children who are already reluctant (33 of 36 cannot answer a greeting) will stop volunteering. The teacher should recast: the child says 'I is Sanjoy', the teacher replies naturally, 'Oh, you are Sanjoy - I am the teacher,' and moves on. The child hears the correct form in context.
- Choose one target per week (this week, 'is/are' or the -s ending) and correct only that systematically; ignore the rest at this stage. Correcting everything is correcting nothing.
- Do correct errors that block meaning immediately and simply, so the pair can continue talking.
- Do not grade Block 2. It is practice; marks come from the Block 3 and Block 4 products.
- For Sanjoy, errors are unlikely in the frame; give him a slightly harder frame ('My favourite ___ is ___ because ___') so the task still stretches him.
Principle: accuracy work belongs in a short, focused slot with a known target; fluency work belongs to speaking time, and fluency is what these 36 children lack most.