The "English only" rule is one of the most widely applied policies in ESL speaking classrooms and one of the most empirically contested. The intuition behind it makes sense: if students can switch to their first language whenever speaking English becomes difficult, they will - and they'll miss the productive struggle that generates acquisition.
The research picture is more complicated. Translanguaging - the use of a student's full linguistic repertoire, including their first language, as a resource in the language learning process - is now a well-established concept in applied linguistics, and the evidence for when first-language use helps versus hinders is more nuanced than a blanket prohibition allows.
YapYapGo is a classroom speaking practice tool for ESL and EFL teachers designed around maximising English production time. Here's the research case for when L1 use supports speaking development and when it undermines it.What translanguaging actually means
Translanguaging is not code-switching (unconsciously switching languages) and not translation (converting from one language to another). It refers to the practice of using all available linguistic resources flexibly and purposefully to communicate and learn.
García and Wei (2014), who developed much of the translanguaging framework, argue that multilingual speakers don't have two separate language systems - they have one integrated linguistic repertoire that they draw on differently in different contexts. Forcing artificial separation of these systems in educational settings may not reflect how language actually works in the brain.
This doesn't mean first-language use in ESL classrooms is always productive. It means the question is more specific: which uses of L1 support acquisition, and which ones avoid the productive struggle that acquisition requires?
When L1 use supports speaking development
Meaning negotiation during pair work. When two students speaking the same L1 briefly switch languages to clarify what they mean before returning to English, this is often productive. The L1 is used to establish shared content, which then enables more complex English production. "Wait - do you mean X or Y?" in L1 followed by a longer English elaboration is a reasonable communicative strategy. Planning before speaking tasks. There is research evidence that allowing students to plan in their L1 before producing in L2 results in more complex, fluent English output than either planning in L2 or not planning at all. This suggests that for teacher-assigned preparation time before speaking activities, L1 planning can be permitted without undermining the English production goal. Explaining complex concepts to lower-level students. In multilingual classrooms where students share an L1 and there is a significant level range, allowing a more proficient student to briefly explain a concept to a lower-level student in L1 can unlock participation that wouldn't otherwise be possible. The alternative is the lower-level student sitting out the activity entirely. Processing emotional content. When speaking activities involve personally significant topics - family, identity, significant experiences - some students find it genuinely impossible to discuss in L2 with the emotional authenticity the topic requires. Brief L1 expression of the emotional core, followed by L2 elaboration, is preferable to either emotional detachment or avoidance.When L1 use undermines speaking development
Using L1 to avoid the productive struggle of L2 production. When students switch to L1 not because L2 is genuinely insufficient for the meaning they need to express, but because L2 is harder and L1 is easier, L1 use is avoiding acquisition rather than supporting it. Whole-pair conversation in L1. When both students in a pair discussion switch entirely to L1, English production drops to zero. This is the clearest case where L1 use is counterproductive. L1 as a default rather than a strategic resource. Students who automatically switch to L1 when encountering difficulty are not developing the repair strategies and circumlocution skills that L2 fluency requires. The difficulty of finding the English way to express something is itself productive. During fluency-timed activities. When the goal is developing L2 automaticity under time pressure, L1 use short-circuits the very process the activity is designed to trigger.Tool tip: YapYapGo is designed for English-medium pair speaking practice. Its structured activity formats - timed questions, visible countdown, automatic rotation - create the conditions that make English the path of least resistance during the activity. A classroom countdown timer is particularly useful: students who know they have 90 seconds tend to maximise English production, since switching to L1 and back takes time they'd rather use speaking.
A practical classroom policy
Rather than "English only" or "use whatever works," a more evidence-informed position:
English always during pair activity time. During the structured speaking activity, the goal is English production. L1 use during this phase should be minimal. Make this expectation clear. L1 permitted during preparation time. If students have 60 seconds to think before speaking, they can plan in their L1 if they find it helpful. The production target is English; the thinking process is their own. L1 permitted for meaning clarification, briefly. A quick L1 check between same-L1 partners ("Does this mean X?") is acceptable if it enables better English production immediately after. L1 never as a substitute for attempting English. The key question students should ask themselves: "Am I using L1 because I genuinely can't express this in English, or because it's easier?" The first is a strategic resource. The second is avoidance.Communicate this policy explicitly rather than just enforcing "English only." Students who understand the reasoning are more likely to make principled choices about language use.
In heterogeneous classes (multiple L1s)
In multilingual classrooms where students don't share an L1, the translanguaging question changes entirely. L1 use in pair work is simply not an option - students have no shared L1 to use. English becomes the default not because of a rule but because it's the only shared medium.
This is one reason why strategically creating mixed-L1 pairs in multilingual classrooms produces more English output than pairing students who share an L1. Not because shared-L1 pairs shouldn't pair together, but because the communication necessity of English is structurally higher in mixed-L1 pairs. See our post on random grouping vs teacher-assigned pairs for more on this.
A random student picker used for pair rotation naturally creates mixed-L1 pairings over time. A classroom group maker is useful when you want to deliberately form groups by L1 background - either mixing or grouping same-L1 students depending on the activity goal.
Sources:
- García, O. & Wei, L. (2014). Translanguaging: Language, Bilingualism and Education. Palgrave Macmillan. - The foundational framework for translanguaging in educational contexts.
- Antón, M. & DiCamilla, F. (1998). Socio-cognitive Functions of L1 Collaborative Interaction. The Modern Language Journal. - Evidence for productive L1 use during planning in L2 tasks.
- Storch, N. & Wigglesworth, G. (2003). Is There a Role for the Use of the L1 in an L2 Setting? TESOL Quarterly. - Empirical investigation of L1 use in L2 pair work tasks.
- Swain, M. & Lapkin, S. (2000). Task-Based Second Language Learning. Language Teaching Research. - The conditions under which L1 use helps versus hinders L2 production.