Comprehensible Input: The 1,500-Hour Bill Nobody Shows You
Updated August 14, 2026 · Reviewed by the Audaz editorial team
Comprehensible input is the idea that you acquire a language by understanding messages, not by studying rules. Listen to Spanish you can mostly follow, in large amounts, and the grammar arrives on its own. It works. The part nobody prints on the poster is the bill. The most popular roadmap for the method asks for 600 hours of listening before you can understand a native speaking normally, and 1,500 hours before you are effective for all practical purposes. The US State Department’s own language school gets students to professional working proficiency in Spanish in 600 to 750 class hours, using the grammar-and-speaking classroom the input crowd tells you to skip. Two roads, roughly the same fare. Input is not the shortcut. It is a different trade.
- Comprehensible input means listening to Spanish you understand about 80 to 90 percent of, in volume, for a long time.
- The Dreaming Spanish roadmap puts level 5 at 600 hours and level 7 at 1,500 hours of input.
- The Foreign Service Institute puts Spanish at 24 to 30 weeks, or 600 to 750 class hours, to reach B2/C1.
- The claim that speaking early “invariably” ruins your pronunciation is the method’s own wording, not a research finding.
- The warning about reading is the one part with real evidence behind it. The fix is hearing a word before you read it, not banning books for 600 hours.
What is comprehensible input, in plain terms?
It is the claim that you pick up a language the way you picked up your first one. By understanding things.
You listen to Spanish that sits slightly above what you can already do. You follow most of it. The bits you miss get filled in by context, pictures, gestures, or the situation itself. Nobody explains the grammar to you. Over enough hours, your brain builds the system by itself, the same way it built English.
The word “comprehensible” is doing the work. Input you cannot follow is noise. Sticking a Spanish film on in the background while you cook is not comprehensible input, because you are not comprehending anything. The method asks for content you genuinely understand around 80 to 90 percent of, so the unknown 10 to 20 percent has something to hang on.
That is the whole idea. It is simple, and the simplicity is a big part of why it spread.
Where did the idea come from?
From one linguist, in the early 1980s, with a very confident claim.
Stephen Krashen argued that people acquire language in only one way: by understanding input that sits just beyond their current level. He called it i+1. Your current competence is i, and the useful input is one step past it. It formed part of a larger set of proposals called the Monitor Model, and it moved language teaching away from grammar drills and towards meaning.
That shift was real and it was healthy. Anyone who sat through years of conjugation tables and still could not order a coffee knows why.
The modern comprehensible-input movement runs on Krashen’s frame. Dreaming Spanish, the biggest Spanish-language version of it, is built on it directly, and it has a published roadmap that turns the theory into hour counts. That roadmap is the most useful document in the whole debate, because it is the only place the method states its own price.
How many hours does comprehensible input actually take?
Here is the roadmap, in its own numbers. Seven levels, each with a stated hour count of input.
The roadmap is honest that these are estimates. It says so on the page. It also gives sensible adjustments: halve the hours if you already speak a close relative like Italian or French, double them for something unrelated like Arabic or Mandarin.
Now put it next to the other well-known number in language learning.
The Foreign Service Institute is the US State Department’s language school. It has been training diplomats for about seventy-five years, and it publishes how long each language takes. Spanish sits in Category I, the easiest group for English speakers. Category I takes roughly 24 to 30 weeks, or 600 to 750 class hours, to reach what the US government calls Professional Working Proficiency. That is Speaking-3 and Reading-3 on the ILR scale, roughly B2 to C1 in European terms.
| Route | Hours | What you can do at the end | How you spend the hours |
|---|---|---|---|
| Foreign Service Institute | 600 to 750 class hours | B2/C1. Work in the language. | Small-group classes, speaking from week one, explicit grammar, homework on top |
| Dreaming Spanish roadmap, level 5 | 600 hours of input | You understand a native speaking normally to you. Speaking still feels like being a child. | Listening only. Speaking and reading discouraged. |
| Dreaming Spanish roadmap, level 7 | 1,500 hours of input | Effective for all practical purposes. | Listening, plus speaking and reading once you are past level 5 |
This is not a fair fight in either direction, and it is worth saying why.
FSI hours are intensive, supervised, full-time, and selected. Students are motivated adults screened for aptitude, taught in tiny groups, with homework on top of the class hours. Almost nobody reading this has that. Input hours are hours you can spend on a bus.
But the comparison still does something useful. It kills the idea that comprehensible input is faster. The method’s own numbers put a functional level of Spanish at 600 hours minimum and full effectiveness at 1,500. That is not a shortcut around the work. It is the same work, arranged differently, with different things bought at the end.
One hour every single day, with no days off, is a bit over four years. Two hours a day gets you there in about two years. This is not an argument against the method. It is an argument against anyone who sells it as effortless. Every honest route to Spanish is measured in hundreds of hours, and the good news buried in that is that the hours are the only requirement. Nothing about you needs to be special.
Does the research actually back it?
Partly. Input matters enormously. The strong version of the claim has never held up.
The oldest objection is the sharpest, and it is now more than forty years old. In 1984 Kevin Gregg pointed out that Krashen “offered no mechanism for converting comprehended input into acquired competence”. You understand a sentence. Then what? The theory says acquisition happens. It does not say how, and it does not say what would count as proof that it did not.
Critics have also noted for decades that Krashen never gave a precise definition of comprehensible input, or a way to identify what i+1 is for a particular learner on a particular day. If you cannot measure the input, you cannot test the claim.
The newer criticism comes from brain imaging. A 2025 conceptual analysis in Frontiers in Psychology by Quang Nhat Nguyen and Dung Thi Hue Doan argues that “understanding input alone, without use or contextual grounding, is not sufficient to rewire the brain’s linguistic capacities”. They point to studies showing that producing language activates right-hemisphere frontal and temporal regions tied to motor planning, which comprehension does not. Speaking engages broader circuitry than listening. They also cite work in which learners who practised conversation with feedback showed greater increases in brain activity than learners who spent the same amount of time passively watching videos in the target language.
Same clock. Different results. That is the finding that should give the strict input position pause.
Nobody serious disputes that large amounts of understood input are necessary. The dispute is whether input is sufficient, and whether the hours spent avoiding speech are hours well spent. Krashen’s ideas are still inside modern research, often under different names. It is the exclusivity that failed, never the emphasis.
Why does the method tell you not to speak?
Because of a specific prediction about damage. It is worth quoting rather than paraphrasing.
At level 4, roughly 300 hours in, the roadmap says that speaking this early “will invariably result in hard-to-fix non-native pronunciation, noticeably bad grammar, and poor word usage”. At level 5, 600 hours in, it softens: purists chasing a near-native accent may want to hold off a little longer, but if you start speaking by then “it’s not a big deal”.
Read the word invariably again. That is a claim about every learner, without exception, and it is presented without a citation. There is a reasonable case that early speaking can fossilise errors, and experienced teachers do see it. But “invariably” is not what the evidence supports, and it is a heavy thing to tell a beginner who is about to move to Madrid.
The other cost is rarely mentioned. Silence is not free. The learners who quit are usually the ones who never got to use the language with a human being. Six hundred hours is a long time to hold your breath, and a learner who abandons Spanish at hour 200 with a clean accent has nothing at all.
If your problem is that you can read Spanish fine but the spoken version washes over you, that is a genuine input problem, and we wrote about why you understand written Spanish but not spoken Spanish separately. If your problem is that you build every sentence in English first, that is a different bottleneck again.
And why does it tell you not to read?
This is where we have to be fair to the method, because this part is not made up.
The roadmap says reading is not recommended until later, “especially if you care about having clear pronunciation”, and keeps saying it through level 5. Sounds absurd for a language as phonetic as Spanish. It is not absurd. There is a real research literature behind it.
Benedetta Bassetti has spent years documenting how written forms leak into second-language pronunciation. Her conclusion is direct: L2 orthographic input “leads to some non-targetlike pronunciations that would not occur if learners were only exposed to acoustic input”. The effects show up after very little exposure to the written form, and they are persistent.
Her examples are precise enough to be convincing. Italian learners of Chinese produce vowel sequences that do not exist in the Chinese they hear, and drop the loudest, longest vowel in the syllable, which is the last one they would drop if they were only listening. Italian speakers of English lengthen a consonant when the spelling doubles it, saying a short t in city and a long one in kitty, because Italian spelling works that way.
English speakers do exactly this in Spanish. Every reader who says tor-TEE-luh learned it from the page, never from a Spanish speaker’s mouth. We took the food version of this apart in our guide to the 15 Spanish food words you say wrong, and the pattern held: the mistakes came from spelling, not from mishearing.
So the diagnosis is right. The prescription is where we part company.
Banning reading for 600 hours throws out a method with a large evidence base of its own. A 2018 meta-analysis of 21 studies covering 1,268 learners found a significant effect of extensive reading on vocabulary. A 2025 meta-analysis in Educational Psychology Review found positive effects across reading comprehension, vocabulary, decoding and fluency, motivation, writing, oral proficiency and general proficiency, with effect sizes from small to medium. Adults benefited most, because they bring more background knowledge to the page.
That 2025 analysis also produced the finding we care about most here. The effects were larger when learners’ choice of text was limited, and when some form of accountability was built in.
“Read whatever you like, at your own pace, with no pressure” is the standard advice, and the meta-analysis suggests it is the weaker version. Constrained choice and a bit of accountability beat total freedom. Which is a long way of saying that a rated list and a deadline work better than a library card. That is the whole design of a reading plan, and it is why we rate books by level instead of handing you a hundred titles.
The honest fix for the pronunciation problem is not silence. It is order. Hear the word before you read it. If you meet a word on a page first, look it up in a native recording before it sets. That handles the orthography leak without costing you the vocabulary that reading brings, and researchers have been testing whether teaching the letter-to-sound rules directly reduces the effect for exactly this reason.
So what should you actually do?
Take the method’s strongest parts. Leave the absolutes.
If books are where you want your hours to go, start with our guide to reading books in Spanish as a beginner.
The verdict
Comprehensible input is the best available answer to the question “what should I spend my time on?” The answer is: understanding Spanish, in volume, for hundreds of hours. That is right, and most learners are not doing it.
What it is not is faster. Its own roadmap says 1,500 hours, and the US government reaches a working level in 600 to 750. Anyone selling input as the shortcut is selling something the roadmap does not claim.
And the two prohibitions do not survive contact with the evidence. The speaking ban rests on a word, invariably, that no study supports. The reading ban rests on a real effect and then overcorrects for it, throwing away a method with decades of meta-analysis behind it to fix a problem that a native audio clip solves in ten seconds.
Listen more than you think you should. Speak sooner than they tell you to. Read the whole time, with your ears open first.
Frequently asked questions
What is comprehensible input in language learning?
It is the theory that you acquire a language by understanding messages, rather than by studying its rules. The input has to be understandable, which usually means content you follow around 80 to 90 percent of. Stephen Krashen proposed it in the early 1980s and called the useful level i+1, meaning one step past your current ability.
How many hours of comprehensible input do you need for Spanish?
The Dreaming Spanish roadmap sets seven levels: 50 hours for common words, 150 to follow learner-adapted content, 300 to understand a patient speaker, 600 to understand a native speaking normally, 1,000 to be comfortable in daily conversation and 1,500 to use the language effectively for all practical purposes. The roadmap calls these general estimates and says to halve them if you already speak a close relative like French or Italian.
Is comprehensible input faster than classroom Spanish?
No, and the method does not claim it is. Its own roadmap puts full effectiveness at 1,500 hours. The US Foreign Service Institute reaches Professional Working Proficiency in Spanish, roughly B2 to C1, in 600 to 750 class hours. FSI hours are intensive and supervised, so it is not a like-for-like comparison, but there is no version of the maths in which input is a shortcut.
Should I really not speak Spanish until 600 hours?
That is the method’s recommendation, not a research finding. The roadmap says speaking before roughly 600 hours “will invariably result in hard-to-fix non-native pronunciation, noticeably bad grammar, and poor word usage”, and it gives no citation for that. There is a reasonable case that very early speaking can lock in errors. There is also strong evidence that production and feedback drive learning in ways listening alone does not. Speaking simply, early, and often is a defensible choice.
Does reading Spanish really hurt your pronunciation?
Written forms genuinely do affect second-language pronunciation. Benedetta Bassetti’s research documents pronunciations that appear only in learners exposed to spelling and would not occur from listening alone, and these effects appear early and persist. The practical fix is order rather than avoidance. Hear a word in a native recording before you read it, and check unfamiliar words you meet on the page. That protects your accent without giving up what reading does for vocabulary.
Does Krashen’s input hypothesis have scientific support?
It has partial support. Nobody disputes that large amounts of understood input are necessary. The disputed part is whether input is sufficient on its own. As far back as 1984 Kevin Gregg noted that Krashen offered no mechanism for turning comprehended input into competence, and critics have long argued that i+1 is too vague to test. A 2025 analysis in Frontiers in Psychology argues from brain imaging that input without use is not enough to rewire linguistic capacity.
Sources
- The Language Learning Roadmap by Dreaming Spanish, v1.3, January 2025, for every level, hour count, known-word estimate, and the quoted guidance on speaking and reading.
- Nguyen, Q. N. and Doan, D. T. H., “Beyond comprehensible input: a neuro-ecological critique of Krashen’s hypothesis in language education”, Frontiers in Psychology, 17 October 2025, for the neurolinguistic critique and for the Gregg (1984) quotation.
- Bassetti, B., “Orthographic input and second language phonology”, in Piske and Young-Scholten (eds.), Input Matters in SLA, Multilingual Matters, 2008, and her later work on orthographic effects in L2 phonetics and phonology.
- Foreign Service Institute language training, for the Category I classification of Spanish at 24 to 30 weeks or 600 to 750 class hours to Professional Working Proficiency (ILR S-3/R-3).
- “Learning a Language Through Reading: A Meta-analysis of Studies on the Effects of Extensive Reading on Second and Foreign Language Learning”, Educational Psychology Review, 2025, for the outcome domains, the small-to-medium effect sizes, and the moderating effect of limited text choice and accountability.
- “The Effects of Extensive Reading on English Vocabulary Learning: A Meta-Analysis”, English Language Teaching, 2018, for the 21 studies and 1,268 learners.
Input is the method. A book is where the hours go. The free Spanish Reading Starter Pack gives you level-rated picks and a plan for finishing your first book in Spanish, so your hours turn into pages instead of streaks. Start with Audaz.
Share
Topics
audazrevista
You may also like
Get the Inside Scoop
