They can read it but they cannot answer it
An English teacher takes Class IV, 42 children, twice a week for 35 minutes. Last week's lesson was a five-line passage about a boy who loses a kite. The teacher read it aloud, the class repeated each line three times, and nearly every child could then read the whole passage accurately. When the teacher asked, "Why was the boy sad?", thirty-one children repeated the last sentence of the passage word for word; nine answered correctly. On the same page there is a picture of children on a roof and a box headed "Let's talk", which the teacher has never used because there is no time. The teacher's plan is to translate every line into Bengali first and then ask the questions again, and to give the class ten more question-and-answer exercises from the workbook. The half-yearly paper carries 60 per cent of its marks on comprehension of unseen passages of this length. The teacher has no supplementary readers, no library period, and no photocopying allowance.
1. Why will translating each line and re-asking the same questions not solve this problem?
Read the worked answer
The failure is at the level of the text, not the word. Thirty-one children answered a "why" question by repeating the last sentence: that is the signature of a reader who has decoded the words but holds no mental model of what happened. Translation gives them a Bengali mental model. They will then answer correctly, and the teacher will conclude that the passage has been understood, when in fact the Bengali has been understood.
Three specific gaps are being missed:
- Lexical depth. The children may know sad as a label but not link lost, kite, string, roof, wind into a causal chain. A word known as a translation equivalent is not yet known as a meaning.
- Inference. "Why was he sad?" is not printed anywhere. The child must connect his kite flew away with he cannot get it back. That move has to be taught, not assumed.
- Question demand. Thirty-one children can locate a sentence; they cannot yet choose between two possible explanations, which is the easier inferential step and must come first.
Translation also consumes the 35 minutes with teacher talk, leaving no time for children to read. More workbook question-answer items repeat the same failed task ten more times, which strengthens the search-and-copy strategy rather than replacing it.
2. Design a 12-minute post-reading sequence using only the textbook, moving from literal to inferential.
Read the worked answer
Keep the reading short and the thinking long. Total 12 minutes.
- 0-2 min: pre-teach four words. lost, kite, string, roof. Picture, gesture, one Bengali gloss each. Four is the limit; more turns the passage into a vocabulary list.
- 2-5 min: read. The teacher reads once at normal speed while children follow with a finger; then pairs read to each other, one line each. Both partners read. No round-robin reading in front of 42 children.
- 5-8 min: Task 1, literal, everyone succeeds. "Which line tells us where the boy was?" Children point, then read the line aloud.
- 8-10 min: Task 2, supported inference. "The boy was happy at the end. True or false? Show me the line." The justification is the assessed part, not the true/false.
- 10-12 min: Task 3, choice plus frame. "Why was the boy sad? (a) His kite flew away. (b) He was hungry." Answer with the frame: "He was sad because ______." Two pairs read theirs aloud.
The picture on the page can be used in Task 1 as the matching target, which costs nothing and gives the weaker readers a second route into meaning.
3. How do you give feedback to the thirty-one children who repeated a sentence instead of answering?
Read the worked answer
Treat the answer as information about the reader, not as a wrong answer. The child did something sensible: found a sentence about the boy. The teaching point is where that strategy stops working, and it can be made once, publicly, without naming anyone.
In practice:
- Accept the sentence as a clue. "Yes, that is the line. It tells us what happened. The question asked why. What does that line make you think?"
- Make the question smaller, not louder. "Was the boy happy or sad at the end?" then "What made him sad?" Two supported steps succeed where one large jump fails.
- Name the move once, on the board. Put the two questions side by side: "Where was the boy?" (answer is in the text) and "Why was he sad?" (answer is not in the text). This is the lesson's actual objective.
- Do not say "wrong" to thirty-one children. Say "that answers a different question", and show the difference.
- Record the pattern as a class objective for the next two weeks. It is not an individual's weakness; changing the plan changes the outcome. If a single child is still stuck after a month of supported inference, that is the point at which individual attention is warranted.