English Code-Switching in a French 2 Information-Gap Task
At a suburban Michigan high school, a French 2 teacher has 28 students in a 50-minute class. During a paired information-gap task about ordering at a café, recordings of six randomly selected pairs show an average of 3.2 English turns per two-minute exchange. Most switches occur when a student lacks a word-straw, spicy, to pay-and the partner answers in English before returning to French. The department policy says « French only, » and a colleague recommends deducting participation points for each English word. The teacher has three weeks before the district interpersonal speaking benchmark, which uses an ACTFL-style rubric and counts 20 percent of the semester grade. Students can sustain four to six turns on familiar topics but rarely use circumlocution or ask for clarification in French. The teacher wonders whether to enforce the policy, redesign the task, or teach repair strategies first.
1. What is the most probable explanation for the code-switching, and what does it imply about the task design?
Read the worked answer
The pattern-3.2 switches per two minutes, clustered at lexical gaps rather than at turn boundaries-points to a vocabulary and strategic-competence problem, not defiance. At Intermediate Low, learners can sustain four to six turns when the lexicon is familiar; the café task introduces low-frequency items such as straw, spicy, and to pay that were not pre-taught, so the information gap is wider than the language gap. English is the efficient repair. A policy that deducts points for each English word would punish the symptom and likely reduce willingness to communicate; it would not supply the missing strategies. The task also may not require French negotiation: if partners can complete the gap in English, they will. A better inference is that the task needs (a) pre-taught lexical chunks, (b) an explicit French repair repertoire, and (c) a reason to stay in French, such as information that only the partner has. The colleague’s recommendation is understandable but misaligned with the developmental level.
2. Outline a three-day instructional sequence to reduce English use without lowering expectations for accuracy.
Read the worked answer
Day 1 should make repair strategies visible. The teacher models a think-aloud while describing a familiar object without naming it, using no more than eight phrases such as « C’est la chose qui… », « C’est le contraire de… », and « Comment dit-on… ? » Students categorize the phrases by function and then practice two in a low-stakes pair task. Day 2 should constrain the temptation to switch: pairs receive « mots interdits » cards and must describe a café item while the partner guesses. The teacher circulates, records two or three strong examples, and debriefs on what allowed the pair to stay in French. Day 3 should repeat the original café task with the same pairs and compare the number of English turns. The point is not to script every utterance but to make French repair the default. Accuracy expectations should remain on the rubric, while the task provides the language needed to meet them.
3. How should the teacher score the district benchmark fairly, given the department’s French-only policy?
Read the worked answer
The benchmark should separate language control from strategic competence so that one weakness does not double-penalize the other. An analytic rubric with four dimensions works: task completion, comprehensibility, vocabulary and language control, and repair strategies. English use should be noted as evidence about strategic competence and vocabulary range, not treated as an automatic zero. If a student says « I need a straw » and then continues in French, the student has communicated but has not used a French repair strategy; that is a rubric-level distinction. The teacher should collect more than one sample, because one two-minute exchange is unstable. Students should see the rubric and self-assess one recording before the summative task. The department policy can be honored as a goal-French is the operating language-without turning the benchmark into a compliance count. This approach is fairer and gives the teacher instructional information.