The Moment You Know You've Lost Them
You've set up the task. The instructions are in the target language. The class is supposed to be discussing in pairs. And then you hear it — English, drifting across the room like smoke. One pair, then two, then half the class. By the time you circle back to the first pair, they're fully, fluently, happily speaking English about the topic you assigned them to discuss in Spanish.
So you remind them. "En español, por favor." They switch back. For about thirty seconds. And then English again.
If you've tried the "no English" rule, you know exactly how this goes. You enforce it. Students find workarounds. They go quiet instead of speaking English. Or they speak quietly enough that you can't hear. Or they comply in your presence and switch the moment you walk away. The rule doesn't change the underlying condition. It just creates performance of compliance.
Why Students Default to English (It's Not Laziness)
The instinct is to frame English-switching as a motivation problem. They don't care enough. They're taking the easy way out. But the SLA research tells a more specific and more useful story.
When learners are given communicative tasks that exceed their current productive capacity — what Krashen's i+1 describes as language pushed beyond the edge of acquisition — they will naturally reach for whatever tool gets the job done. For students in a secondary school classroom, that tool is English. It's not a moral choice. It's a cognitive one. The communicative demand exceeds the available resources, and they solve the problem with what they have.
Swain's output hypothesis is useful here too. The push to produce is necessary — students need to attempt output to notice the gaps in their interlanguage. But "attempting output" and "being assigned a task in the target language" are not the same thing. A task that's too open, too abstract, or too far beyond current proficiency doesn't produce pushed output. It produces English.
There's also the social dimension, which the research on affective filter captures but classroom experience fills in. Talking with a peer is fundamentally different from talking to a teacher. With a peer, the stakes of sounding incompetent feel higher, not lower. Students would rather sound fluent in English than incompetent in Spanish with someone they sit next to every day. The social cost of imperfect L2 production between peers is real, and no amount of rule-enforcement changes the calculation.
What Doesn't Work (And Why)
The no-English policy fails for a simple reason: it addresses the symptom, not the condition. The condition is that students don't have enough compelling reason, cognitive support, or lowered social stakes to produce in L2. Telling them the rule doesn't change any of those three things.
Pair work with open-ended prompts is the second biggest culprit. "Discuss the theme of family in the story" is a great prompt for an advanced composition class. For a second-semester Spanish student, it's an invitation to switch to English within eight seconds. The task is too cognitively demanding and too socially exposed.
Points for target language use — a grade penalty for English — works only insofar as it pressures compliance. And compliance and acquisition are different things. A student who speaks broken Spanish anxiously under threat of a grade deduction is not acquiring. Their affective filter is spiking, which is precisely what blocks acquisition. You've produced L2 performance at the cost of the conditions that make L2 acquisition possible.
The Film Intervention That Actually Shifts the Condition
Here's what changes when a film is in the room. The shared reference.
When students are asked to discuss something they both just watched, the communicative task narrows dramatically. They're not generating ideas from scratch — they're responding to something specific that happened on screen. "What did he mean when he said that?" has a referent. The answer isn't constructed from abstract vocabulary and grammar; it's anchored to a scene, a face, a moment of tension or humor or surprise that both students just experienced together.
That anchoring does something important for code-switching. When the task is manageable — when students have the vocabulary because they've just heard it in context, when the referent is specific and shared — the cognitive load drops enough that L2 becomes viable. They don't switch to English because they don't need to. The task is tractable in the target language.
The emotional element matters too. Students who are genuinely curious about what just happened on screen have a motivation to discuss it that transcends the grade. Curiosity is an intrinsic communicative drive, and intrinsic drives produce language that compliance never does. When a scene is genuinely surprising, or funny, or disturbing, students want to say something about it. Give them the scaffolding to say it in the target language, and they will.
Structural Fixes That Reduce Switching
Make the Task Smaller and More Specific
The more specific the task, the more manageable it is in L2. "Describe what happened in the kitchen scene using three details" is easier than "Discuss the relationship between the two characters." Specific tasks have specific vocabulary. Vague tasks require the full generative capacity of a language students are still acquiring.
Pre-Load the Vocabulary Before the Task
Students switch to English when they hit a lexical gap. If they encounter a moment in a scene and don't have the word for what they just saw, they'll say it in English, and once they've switched, they stay switched. Pre-loading key vocabulary — not a list, but the specific words they'll need for the specific task — reduces the frequency of those gaps enough to change the pattern.
Use Written Scaffolds as a Bridge
Having students write one sentence in the target language before speaking in pairs changes the dynamics significantly. Writing is private. The fear of sounding incompetent in front of a peer is absent. Students will attempt language in writing that they won't attempt in speech. Once they've written it, speaking it becomes retrieval rather than generation. The cognitive load drops, and so does the incentive to switch.
Reward the Attempt, Not the Accuracy
The research on affective filter is unambiguous: when learners fear evaluation, acquisition stops. If students believe that switching to English is less bad than producing imperfect Spanish, they will switch. Changing that calculation means making imperfect L2 production genuinely safer than English. Call on the student who produced a broken but target-language sentence and respond to the content, not the form. That's the signal that changes the room.
The Deeper Problem
When a class keeps switching to English, the honest diagnostic question is: is the input they're getting comprehensible and compelling? Krashen's argument is that when those two conditions are met, acquisition happens and the motivation to produce in L2 follows. When classes are stuck in English, it's usually because one or both conditions are missing.
Film addresses both simultaneously. The visual context makes input comprehensible beyond what the language alone would allow. The narrative drive — the genuine interest in what happens next — makes it compelling. A well-chosen scene in the target language, with the right before and after structure, creates the conditions where English stops being the default because the target language stops feeling impossible.
That's not magic. It's design. And the design is what you're building when you put a structured film unit in front of a class instead of a conversation prompt and a hope.
If you want to see how the during-film and post-film tasks in a structured lesson plan create exactly that architecture, the ESL lesson plans are a good place to start — this particular problem tends to be acute in classes where English is the dominant ambient language outside the classroom.
