Repetition Without Consequence: Banding the Effect Marks in a Poetry Item
Four examiners are finalising the scheme for the Poetry section of a Bodo First Language terminal paper. Question 6 carries 5 marks: 'Identify the figure of speech used in the repeated line of the prescribed poem and explain what the repetition contributes to the poem's mood.' The repeated line addresses a river. The draft scheme gives 2 marks for naming the device and 3 for the explanation. In a trial batch of 20 scripts, 17 candidates named repetition correctly, but the explanations fell into three clusters: (i) 'the poet repeats it because it is the main line' - 9 scripts; (ii) 'repetition makes the poem musical and gives pleasure' - 6 scripts; (iii) 'each repetition returns to the river after a new image, so the mood turns from complaint to acceptance' - 2 scripts. Two examiners want full explanation marks for cluster (ii); a third calls it a stock phrase that could be pasted into any poem. The paper must be finalised the same afternoon; the section totals 20 marks.
1. Distinguish the three explanation clusters for the 3 explanation marks and propose a defensible split.
Read the worked answer
Separate the two constructs. Identification (2) is mechanical - 'repetition' or its Bodo equivalent earns it; 17 of 20 did. The explanation marks must test whether a candidate can link the device to this poem's movement.
Apply a transferability test.
- (iii) names the structural work the repetition does: each return follows a new image, and the mood shifts from complaint to acceptance. Full 3.
- (ii) 'musical and gives pleasure' fits any poem. If the candidate names the sound effect (the line as a refrain marking the stanzas), give 1; with no such anchor, still 1.
- (i) restates the line instead of explaining it. 0.
Why cluster (ii) must not get 3. That would turn the explanation mark into a second identification mark, and 15 of 20 scripts would sit at the top; the item would stop discriminating. Band the explanation 0/1/3 and give markers the anchor question: could this sentence be pasted into another poem?
2. One examiner proposes a scheme note that Devanagari spelling variation (anusvara versus conjunct) should cost marks in Question 6. Rule on this.
Read the worked answer
Refuse the note. Question 6 assesses response to a poem. Devanagari spelling habits are a different construct and are not part of the 5-mark split. Deducting here would make the poetry total partly an orthography score, would penalise candidates for schooling or handwriting, and would tell nobody what the candidate can do with a poem.
Where to draw the line. If a variant makes a word unreadable, the general legibility instruction applies: award what can be read. If spelling is so unstable that meaning is in doubt, ask a second reader rather than deducting unilaterally.
If spelling is to be assessed at all, it needs its own item with a stated scheme, so candidate and marker both know what is tested. Record the ruling in the scheme so all four examiners apply it, and check that no marker has already begun penalising.
3. A candidate names a different device (a simile) but the explanation accurately describes what the repetition does. How do you mark it?
Read the worked answer
Mark the parts separately.
- If the named device is genuinely present in the line and the explanation describes its real effect, credit both parts: a poem can carry repetition and simile, and the key is not exhaustive.
- If the device is misnamed (no comparison in the line) but the explanation accurately describes the repetition's work, the 2 identification marks are lost; the explanation can still earn marks, because it demonstrates the reading skill the item tests.
- If the device is misnamed and the explanation is generic, the script scores 0–1.
Annotation and consistency. Note which device was named on each script, so the second examiner sees why identification was refused. Before marking, agree the list of devices actually present in the line, so the 'also present' argument is settled once rather than script by script.