Fossilised measure-word omission and 了 overuse in an advanced learner

The patterns look like a stable interlanguage system rather than random slips. Measure-word omission is a plausible simplification: the L1 has no classifier system, and omission rarely blocks understanding in context, so the learner receives little negative evidence. The overgeneralised 了 reflects a one-to-one mapping strategy: past time is marked by 了, which works often enough to be reinforced. Avoidance of 把 is a communication strategy: the topic-comment alternative is grammatical and sufficient, so the learner never has to process the construction under pressure.

Fossilisation is sustained by three conditions. First, communicative success: her clients understand her, so there is no push to change. Second, insufficient salience: measure words are unstressed and often swallowed in native speech, so she may not notice them in input. Third, feedback that is immediate but not processed: repetition after correction shows she can imitate, not that she has noticed the rule. The previous course's drills may have raised affective filters without creating awareness of when the forms are obligatory. The goal is therefore not more correction but structured noticing plus pushed output in tasks where the target form is needed for accuracy, not just meaning.

Each lesson follows a cycle: a short awareness task, a focused production task, and a transfer task. Week 1 targets measure words through a listening and sorting task: she listens to six short client voicemails and marks which noun phrases need a measure word, then re-records two messages using 这/那 plus a measure word. Week 2 targets 了 through a timeline and contrast task: she sorts past events into completed-action 了, change-of-state 了, and no 了, using her own work week, then explains three completed actions on the phone. Week 3 introduces 把 through a problem-solving task: rearranging a delivery, moving items, returning goods, where 把 is the natural choice; she compares her topic-comment version with a 把 version and identifies the difference in focus. Week 4 integrates all three in a simulated client call with a checklist.

This is focus on form, not drills: the forms arise from meaningful tasks, but the teacher briefly highlights the rule and gives corrective feedback on the target only. The tasks are adult and work-related, which addresses her stated goal of sounding professional. Each week recycles earlier targets, because one encounter will not destabilise a fossilised form.

Measurement should use the same task type before and after: a recorded 5-minute client call role-play, scored for obligatorily supplied measure words, appropriate use of 了, and appropriate use of 把, plus a short delayed task two weeks later to check retention. The teacher should count obligatory contexts, not total errors, so the measure is comparable across performances. She should also ask the learner to self-correct a transcript, which taps awareness rather than imitation.

If the learner still reverts, the teacher should not simply repeat the same four weeks. First, check whether the target was too broad: maybe measure words in 这/那 frames should be consolidated before adding 把. Second, increase the salience of input: use short recordings where the target is prominent, and ask the learner to notice rather than produce. Third, negotiate the goal: if 把 remains avoided but the learner can use it in written client messages, that is progress. Fourth, consider that one hour a week is a low dose; the teacher should build two 5-minute asynchronous tasks between lessons. The professional judgement is to distinguish a slow rate of change from a genuine ceiling and to adjust the plan on evidence rather than blame the learner.