Seven red circles in a five-sentence paragraph
Class V English, 34 children whose first language is Malayalam, four periods a week of 40 minutes each. The teacher sets a weekly dictation and a five-sentence paragraph on a familiar topic, and enters the marks in the register. Fathima's paragraph this week reads: 'Yesterday I go to my grandmother house. She give me a mango. I am very happy.' The teacher circles seven errors in red, writes 'Careless - check tenses', and gives 4/10. Last month Fathima wrote eight sentences and made eleven errors; she now writes two short sentences and checks them repeatedly before handing in. In class she rarely volunteers; when called on she answers in Malayalam and looks at the floor. There is no oral English routine beyond the teacher reading the textbook aloud and the children repeating after her. Most of the class shows the same kinds of errors, and six children have stopped writing more than three sentences.
1. Analyse Fathima's errors. What do they indicate, and why is this feedback likely to slow her learning?
Read the worked answer
The errors are rule-governed and developmental, not careless.
- go for went: the past tense is not yet internalised; irregular forms are learned late and are often overgeneralised before they settle.
- grandmother house: the possessive is absent. Malayalam marks possession with a case suffix (അമ്മൂമ്മയുടെ), not with a separate possessive word, so transfer predicts exactly this omission.
- She give: third-person -s is absent. Malayalam verbs do not agree with the subject in person, so again the L1 gives no support.
Each error is an intelligible hypothesis about how English works. Calling this 'careless' attributes a stage of interlanguage to a character flaw, which the child cannot act on.
Why the feedback slows learning:
- Seven circles on five sentences is more correction than a learner at this stage can process, and none of it is prioritised, so nothing is learnable.
- 4/10 tells her the whole text failed, although the message is perfectly clear - the reader can see the grandmother, the mango, the happiness.
- The behavioural evidence is the strongest signal: output shrank from eight sentences to two, and she now checks repeatedly. That is avoidance, not improvement. Less output means less practice, so progress slows further. She is managing anxiety by playing safe, which is exactly what a heavily penalised second-language writer does.
2. What feedback policy would you adopt for this class for one term, and how would you handle Fathima's paragraph specifically?
Read the worked answer
Policy: one target feature at a time, meaning first.
- Each piece of writing has a single announced target - here, the past tense of five common verbs. Only target errors are counted; other errors get a light dot, noticed but not scored.
- Comments start with content: 'I can see the mango and the happiness - this is clear.'
- Score on a three-criterion rubric: message clear / sentences complete / target feature used. Report criteria separately so that a communicatively strong piece is not wiped out by grammar.
- One revision, not a rewrite. The child changes the target feature only.
- Oral rehearsal before writing: say the sentences to a partner in English, then write what you said.
Applied to Fathima's paragraph: the target is went / gave / was. Give her a word bank of irregular past forms, ask her to find and change only the verbs, and keep 'grandmother house' and the missing -s for a later target. Then ask her to add one more sentence - length is a goal in itself at this stage.
Start a personal error log with her, listing only the current target. When the past tense is secure in speech and writing, move to possessives. Involvement in spotting her own error once per piece matters more than a comprehensive red pen.
3. Design a 10-minute assessment that would show Fathima's progress in a month, and explain what makes it fair for a learner at this stage.
Read the worked answer
Task: a picture of a simple event (a market scene), a word bank of six verbs including three past forms, and 4 minutes to write four sentences, preceded by 3 minutes of oral planning with a partner. Then two oral questions about the picture, answered in English.
Scoring, reported separately:
- Communication: can I understand what happened?
- Length: number of complete sentences (target: from two to four).
- Target feature: correct past forms out of the forms attempted.
- Risk-taking: does she attempt an irregular verb rather than avoiding it?
Why this is fair:
- The topic and vocabulary are known and supplied, so the task measures writing, not guessing.
- Errors outside the current target are not penalised, which matches the term's teaching focus.
- Oral planning uses her stronger modality before the weaker one, which is normal for a learner at this stage.
- It is scored ipsatively - against her own previous sample - as well as against the rubric, because comparing an early L2 writer with fluent classmates measures home language exposure, not learning.
- Half the marks come from speaking, so a child who is orally ahead is not invisible.
Use a parallel picture after a month to test transfer, and give her the result as a description of what she can now do, not only as a number.