Separating rule formation from imitation in a child language recording
Marcus is analysing a 10-minute recording of his niece, aged 2 years 6 months, playing with a parent. The transcript contains: 'I runned to the park'; 'teddy falled down'; 'more juice'; 'no nap, no nap'; 'Mummy sock on'; 'what that is?'; and a stretch where the child repeats 'big red bus' after the parent says it. The parent often expands: child: 'teddy falled'; parent: 'yes, teddy fell down'. Marcus wants to claim that the child 'has learned the past tense rule' and that 'the parent is teaching by modelling'. His essay must evaluate evidence for rule formation versus imitation and consider limitations of a single recording. He has 40 minutes and a 900-word limit.
1. Which utterance is the strongest evidence for rule formation, and which for imitation? Explain the distinction using the data.
Read the worked answer
Strongest evidence for rule formation: I runned to the park and teddy falled down. The child applies the regular past-tense -ed to irregular verbs that adults typically model as ran and fell. If the child were simply imitating, she would be more likely to reproduce ran and fell. Overgeneralisation suggests she has extracted a pattern and is applying it productively, even though the output is non-adult.
Strongest evidence for imitation: the stretch where she repeats big red bus immediately after the parent. Immediate repetition can be imitation, though it may also be practice or participation.
Important nuance: one recording cannot establish that the child has learned the past tense rule in a permanent sense. It shows rule-like behaviour in this session. Marcus should also distinguish overgeneralisation from formulaic chunks: what that is? may be a partially analysed question frame, not evidence of a complete rule. The parent's expansion teddy fell down is a recast, which may support learning, but the data do not prove causation.
2. How should Marcus use the parent's expansions without overclaiming that the parent is 'teaching'?
Read the worked answer
Marcus should describe the parent's expansions as input that may scaffold the child, not as proof of teaching. In the exchange, the child says teddy falled; the parent responds yes, teddy fell down. That is a recast: the parent repeats the child's meaning in a more adult form. Recasts can make the target form salient, but a single session cannot show that the child notices or retains it. Marcus should also note that the parent's expansion is contingent on the child's utterance, which is consistent with child-directed speech that supports conversation.
A defensible claim: 'The parent's recasts provide immediate, contingent models of the irregular past tense, which may support the child's developing system.' An overclaim would be: 'The parent teaches grammar by correcting the child.' The transcript does not show explicit correction, and the child does not repair the form in the extract.
3. Write a short evaluation paragraph that acknowledges one limitation and still makes a defensible claim.
Read the worked answer
Sample evaluation paragraph:
The recording shows productive overgeneralisation in I runned and teddy falled, which is consistent with the child having extracted a past-tense pattern rather than merely repeating adult forms. However, a ten-minute recording cannot establish the stability of that pattern, nor can it show how the parent's recasts affect later production. The data therefore support a limited claim: in this session, the child applies -ed to irregular verbs and receives contingent recasts from the parent. A stronger study would sample multiple sessions, compare the child's forms with the frequency of the input, and test whether the child later repairs runned to ran. Marcus should present the evidence as illustrative, not definitive.