Ask a language teacher "how do you build fluency?" and you'll get one of two answers. Half will say: consume as much Spanish as possible — read, listen, absorb. The other half will say: you have to produce language — speak, write, push yourself to use it.
Both camps are right. The debate, it turns out, is not about which is better. It's about what comes first, in what proportion, and at which stage of learning.
Krashen's Input Hypothesis: the foundational argument
In the early 1980s, linguist Stephen Krashen proposed something radical for the time: that we don't actually "learn" a second language the way we learn maths. Instead, we acquire it — subconsciously — by being exposed to language we can mostly (but not entirely) understand. He called this "comprehensible input" and labelled it i+1: your current level (i) plus just one step ahead (+1)[1].
Krashen's argument was that grammar instruction, output practice, and error correction were essentially irrelevant to true acquisition. What mattered was the quantity and quality of comprehensible input. Speak enough, read enough, listen enough — and the grammar would emerge naturally.
The practical implications were significant. Krashen's ideas directly inspired immersion programmes, sheltered instruction models, and the broad shift away from grammar-translation teaching in the 1980s and 90s. They're also, in some form, behind every app or service that tells you to "consume content at your level."
The Output Hypothesis: Merrill Swain's challenge
Krashen's model had a problem. Immersion students in Canada — children who spent years in French-language schools — consumed enormous amounts of comprehensible input. And yet, after years of immersion, many still produced grammatically incorrect French in ways that didn't reflect what they'd heard.
Merrill Swain, who studied these learners, proposed the Output Hypothesis: producing language — speaking and writing — serves functions that input simply cannot[2]. Specifically, output forces you to notice gaps in your knowledge. When you try to say something and can't, you become aware of what you don't know. That awareness makes subsequent input much more effective.
"Pushed output — being required to say or write something precisely — is a catalyst for noticing the gap between what you can express and what you need to express."— Merrill Swain, 1985
What the research synthesis tells us
Decades of subsequent research — including a landmark meta-analysis by Norris and Ortega (2000) of 49 experimental studies — suggests that explicit instruction combined with meaningful input and structured output practice produces the fastest, most durable gains[3]. Neither input-only nor output-only approaches are optimal.
Nick Ellis's research on "usage-based" language acquisition adds another dimension: fluency is built through repeated encounters with language in meaningful contexts[4]. The more times you encounter a word or structure — across different contexts, in both input and output — the more automatic and retrievable it becomes.
What this means practically — at each stage
The role of vocabulary in the input/output equation
Both input and output are bottlenecked by vocabulary. You can't comprehend input you don't have the words to decode. And you can't produce output for concepts you don't have the words to express.
This is why vocabulary acquisition sits at the centre of both theories. Without a sufficiently large, accessible lexicon, neither input nor output functions effectively. Input becomes noise. Output becomes paraphrase and approximation.
The practical implication: don't treat vocabulary study as separate from input/output practice. The most effective approach is to study vocabulary in context — from real texts you've read or listened to — and then immediately use it in output tasks. For a deeper look at why vocabulary is the actual bottleneck, see our guide: Why vocabulary is the foundation of every language skill →
The takeaway
The input vs output debate has been mostly resolved by research: both matter, sequencing matters, and the right balance shifts with your level. What doesn't change is this — passive consumption alone will not make you fluent. At some point, you have to speak. You have to write. You have to be pushed to the edge of what you can express — and then pushed further.
The discomfort of reaching for a word and not quite finding it? That's not failure. That's acquisition happening. And if you're preparing for a specific exam, how you balance input and output changes significantly — see our guides for DELE B2 vocabulary preparation and IB Spanish vocabulary by topic.