You’re not bad at languages. You were taught backwards.
You learned your first language without a single grammar table — just years of understanding messages aimed at you. That isn’t a charming story. It’s the most robust finding in language science, and it’s the whole bet behind rianthai.
Understand messages.
That’s it. That’s the method.
Comprehensible input is language you can understand— a little bit beyond your current level, but still followable from context, pictures, and tone of voice. The linguist Stephen Krashen called this “i + 1”: your current level, plus one small step.
When you understand a message, your brain quietly absorbs the patterns that carried it — the grammar, the word order, the sounds — without anyone teaching them. Krashen’s claim is blunt: “We acquire language in one way, and only one way: when we get comprehensible input in a low-anxiety environment.”
So the work isn’t studying Thai. The work is understanding Thai— for hundreds of hours, on things you actually want to watch. Speaking and reading grow out of that foundation later, the same order they did when you were small.
Your brain is a pattern-detector. Input is the only thing that feeds it.
This isn’t mysticism. It’s how human cognition handles language, and the evidence for the pieces is some of the sturdiest in the field.
You learn patterns without noticing
Decades of cognitive science show humans extract grammatical regularities from input below the level of awareness. You don’t memorize the rule; you absorb the pattern from a thousand examples.
Comprehension frees up the brain
We’re limited-capacity processors who pay attention to meaning first. When input is understandable, attention is freed to soak up howit was said — which is why it should stay ~95% clear.
It’s how you already did it
You acquired your first language from years of comprehensible input, no grammar drills, with a long listening period before you spoke. The same statistical machinery is still running in your adult head.
of invented words a reader absorbed from a novel with no dictionary and no instruction — acquisition from pure input.
the record of comprehensible-input methods against grammar-first methods in head-to-head comparative studies.
the efficiency Mason & Krashen reported for input — ~70 hours of CI matching ~286 hours of traditional study. (An advocate study — promising, not yet consensus.)
“I studied French for six years and can’t order a coffee.”
It’s not you. Among Americans fluent in a second language, only about 7% learned it in school — fewer than one student in a hundred. Drilled grammar “lives in their heads as something they know, not something they can do” — and it collapses the moment a real conversation starts.
To be fair: the classroom wins where it’s aimed. Explicit grammar study is genuinely better for a grammar exam next month and for measured written accuracy. And gamified apps are great at one thing — getting you to open the app. They’re just not the same thing as acquiring the language; the streak can become its own goal.
What CI is bad at — and what we do about it.
A method worth trusting can say where it’s weak. Here are the real downsides, none of them hidden — each next to how the product answers it.
The start feels slow.
For the first stretch there’s no lesson to “complete” — just hours banking up. So we draw the whole road, honestly, and show the bar moving. You see progress because it’s real, not because we invented a number.
Speaking lags behind understanding.
You’ll understand far more than you can say for a long while. That gap is normal — and for a tonal language it’s protective: your ear sets the target before your mouth commits. The dashboard de-emphasizes output until your listening is ready.
Pure input isn’t the whole story.
Mainstream research is clear that input is necessary but not sufficient — a little scaffolding helps. So we add it in your native language: instant definitions, sentence explanations. We never generate fake Thai for you to practice on. The input stays human.
Good beginner Thai content is scarce.
Spanish learners have thousands of hours ready to watch; Thai doesn’t have that yet. That scarcity is the entire reason this tool exists — it wraps the real Thai content that does exist with transcripts, lookups, and leveling so you never run dry.
It rewards consistency.
There’s no streak mechanic guilting you back at 9pm. Just honest hours, a gentle weekly goal, and zero loss-aversion pressure. The point is time with content, never time in the app.
Aren’t subtitles and flashcards cheating?
Short answer: no — with one rule. Scaffolding helps when it hangs off real human input and stays in your native language; it hurts when it replacesthe input. Here’s exactly where the research draws that line — and why our captions and word-saving are built the way they are.
Captions: Thai on screen, English on hover
The mode that hurts is English subtitles. In a controlled study, an hour of video with native-language subtitles produced a 0% listening gain — worse than no subtitles — because when you read your own language your ears switch off. (The key study is literally titled “native-language subtitles harm foreign speech perception.”)
Same-language Thaicaptions do the opposite — a meta-analysis puts the listening and vocabulary gains both large — and for Thai they pull a special trick: they show you where one word ends and the next begins, the exact skill Thai’s space-free script otherwise hides (and the gain transfers to speakers you’ve never heard).
So we keep Thai on screen and English to a hover — and let you blur each line and listen first, so captions stay a check, not a crutch.
Saved words: fuel for the input, not a detour
Picking a word up from context alone takes roughly 8–10+ encounters, and the meaning is the slowest part to stick. A quick look-up roughly doubles what you retain from the same reading.
And it isn’t a detour from input — it feeds it. You need to know ~95–98% of the words for input to be comprehensible at all, and deliberately learning the high-frequency words you keep meeting is the fastest way there. The field’s standard balanced-program framework, Nation’s “Four Strands,” budgets a full ~25% of a healthy course to exactly this.
The catch: the words must come from your real videos and be reviewed in their real sentences (with the audio), not a decontextualized list — which is why we save the line, not just the word. And flashcards aren’t inert: deliberately-learned words wire into your automatic mental lexicon.
The strictest listening-first camp — ALG, born teaching Thai — says use no subtitles and no look-ups at all early on. We think that’s too absolute (the evidence for permanent “damage” is thin), but they’re right about the failure mode: captions you can’t yet read, or a flashcard habit that eats your watching time, are a step backward. So we hold these back until they help — once you can read a little Thai — and keep hours of input the only number that counts.
Thai is hard in exactly the way input is good at.
Thai’s difficulty isn’t grammar — there are no verb conjugations, no tenses, no genders, no plurals. The hard part is your senses: five tones where ใหม่ ไม้ ไหม้ ไม่ ไหม are five different words, a script with no spaces between words, and almost no cognates to lean on.
Those are perceptual problems — and perception is built by listening, not by rules. A meta-analysis of 25 years of research found that training your ear improves your accent too. Listen first; the mouth follows. (And because Thai script has no word-spaces, reading has to come afteryour ear can already segment speech — not before.)
Here’s the part most people don’t know: this whole movement was born teaching Thai. Dr. J. Marvin Brown built the listening-first method at the AUA school in Bangkok, and Dreaming Spanish’s famous roadmap was “initially based on thousands of students who learned Thai with comprehensible input at AUA.” Spanish is the famous success story. Thai is the original laboratory.
the hours Thai takes vs. Spanish to reach professional proficiency, per the US State Department’s FSI — but your early listening milestones arrive on the standard schedule, because that schedule was built on Thai.
“I first came across Krashen’s idea that we acquire languages by understanding messages, and in no other way.”
— Dr. J. Marvin Brown, who built the method teaching Thai in Bangkok
One honest caveat: Brown’s strongest claim — that speaking too early permanently damages your accent — is an influential observation, not a settled experiment. We lean on what is well-supported: for tones, listening first is simply the wise way in.
Input is the foundation no one acquires a language without.
We’ll be straight with you: comprehensible input isn’t a magic wand, and serious researchers don’t claim it’s the onlything — interaction and speaking matter too, especially for polish. But on the one point that decides everything, the field agrees: you cannot acquire a language without massive understandable exposure to it. That’s the engine. Everything else is a tune-up.
So the whole bet is simple: make that engine — thousands of hours of compelling, comprehensible Thai — as frictionless as humanly possible, and let understanding grow on its own. That’s the most defensible ground in language learning. It’s also, conveniently, the most fun.
Convinced? Now see the road.
Every hour has a place on it, and every milestone is something you’ll actually be able to do. Walk it before you start.
Every claim above links to its source. Here they are gathered — including the studies that complicate the picture, because a page that only cited its cheerleaders wouldn’t deserve your trust.
Foundations
Evidence & effect sizes
The honest counter-case
Thai & tones
Subtitles & vocabulary
- Mitterer & McQueen (2009) — native subtitles harm, foreign help
- Birulés-Muntané & Soto-Faraco (2016) — L2 vs L1 subtitles for listening
- Montero Pérez et al. (2013) — captioning meta-analysis
- Yanagisawa, Webb & Uchihara (2020) — glossing meta-analysis
- Nation — the Four Strands (deliberate study ≈25%)
- Elgort (2011) — deliberate word learning becomes automatic