The Output Hypothesis
Why do good listeners still struggle to say it right?
The Output Hypothesis, proposed by Merrill Swain, holds that producing language, especially when pushed to be precise, helps learners notice gaps, test hypotheses and reflect on form in ways comprehension alone may not.
In Canada’s French immersion programmes, Merrill Swain met a puzzle. Children who had studied school subjects through French for years understood it very well, yet their speaking and writing remained less grammatically accurate than that of francophone peers. If comprehensible input were enough, these learners should have been further along.
Swain (1985) argued that what they lacked was output, and in particular pushed output: production that demands precision, coherence and appropriacy. She called her proposal comprehensible output; it is now usually known as the Output Hypothesis. Her central claim was that speaking or writing pushes learners beyond processing for meaning, which is often enough to understand, to processing grammar, which is needed to say things accurately. Swain (1995) later named three functions of output: noticing gaps in one’s knowledge, testing hypotheses about how the language works, and reflecting on language metalinguistically.
The evidence is real but uneven. Swain and Lapkin (1995) recorded learners thinking aloud while writing and showed them noticing problems and reworking their language. Shehadeh (2002), however, pointed out that much research had documented output rather than shown that it led to acquisition. Krashen (1998) went further, arguing that comprehensible output is too rare to account for much acquisition. Swain’s later work recast output as collaborative dialogue, in which learners solve language problems together.
For teachers, the point is not simply more talk but talk that stretches. Tasks that make learners reformulate, explain or write for a reader push them past the shortcuts meaning-focused talk allows, which is why output pairs naturally with noticing and task-based teaching.
In actionTwo learners rebuild a story
- S1
- The girl buy three tomato at the market.
- S2
- Bought. It’s the past. And three tomatoes.
- S1
- Tomatos or tomatoes? How do you write it?
- S2
- With e, like potatoes. I’m quite sure.
Writing the text together makes both learners notice gaps and test hypotheses aloud, the kind of talk about language that Swain and colleagues call language-related episodes.
In the classroom
- 1
Run a dictogloss: read a short text twice at normal speed, have pairs reconstruct it in writing, then compare their version with the original and discuss the differences.
- 2
Answer some errors with a clarification request (“Sorry, when did that happen?”) instead of reformulating, so the learner has to repair the utterance and produce the form themselves.
- 3
After group discussion, add a planning stage in which groups prepare a careful spoken or written report for the class, moving from informal talk to more precise language.
Myth and reality
Output is only practice: it makes known language more fluent but teaches learners nothing new.
Swain argues output can itself be a site of learning, because trying to say something precisely shows learners what they cannot yet say. How large that contribution is, compared with input, is still debated.
Reflect
In your lessons, when do learners have to say something so precisely that a vague, half-grammatical version would not do?
Worth a minute before your next lesson, or a conversation in the staffroom.
SourcesChecked against Crossref and library records
Swain introduced comprehensible output in 1985, using evidence from Canadian French immersion. She set out three functions of output in 1995 and later reframed output within sociocultural theory as collaborative dialogue and languaging.
Swain, M. (1985). Communicative competence: Some roles of comprehensible input and comprehensible output in its development. In S. M. Gass & C. G. Madden (Eds.), Input in second language acquisition (pp. 235–253). Newbury House.
Swain, M., & Lapkin, S. (1995). Problems in output and the cognitive processes they generate: A step towards second language learning. Applied Linguistics, 16(3), 371–391.
Think-aloud data showing learners noticing problems as they write and working their way to better versions.
Kowal, M., & Swain, M. (1994). Using collaborative language production tasks to promote students’ language awareness. Language Awareness, 3(2), 73–93.
Classroom-facing: teenage learners of French reconstruct a text together, with practical lessons about task design and grouping.
Shehadeh, A. (2002). Comprehensible output, from occurrence to acquisition: An agenda for acquisitional research. Language Learning, 52(3), 597–647.
A careful review arguing that research must show output leads to learning, not just that it happens.
Krashen, S. (1998). Comprehensible output?. System, 26(2), 175–182.
The input camp’s reply: a short, pointed case that output plays at most a minor role in acquisition.
ConnectionsSee it in the graph
- Comprehensible Inputnot yet writtenSwain’s proposal began as a challenge to the sufficiency of input.
- The Noticing HypothesisOutput’s first function is to make learners notice gaps.
- Interaction Hypothesisnot yet writtenNegotiating meaning gives learners both feedback and a push to modify output.
- RecastsRecasts supply the form; prompts like clarification requests push learners to produce it.
Know a teacher who’d like this?
Send them the card. It’s made for Facebook, Zalo, LinkedIn and Instagram, and links back here.