Most people think language learning is a single climb: a steady slope from zero to fluent, where effort is the only variable. Put in the hours, get the result.
The reality is more like a path full of obstacles. Not a language barrier: that's between speakers.
Language learning obstacles are the specific challenges every learner has to work through on the way to fluency. Some are small detours, others are genuine dead ends if you approach them wrong. And critically, each one requires something different to get past. Treating them all the same, by throwing uninformed 'effort' at them, is how most learners end up burned out and stuck.
This article maps the major obstacles by CEFR level. Each transition has its own distinguishing themes, and naming the right one is most of the battle: it unlocks the strategy that actually works. At Atlas Runa, we've broken language learning down into over 250 micro-obstacles, each with its own fix, and built a system around clearing them with the right tool at the right time. Here, we tease out the broad themes per proficiency level: the major reasons why most learners stall at each stage to fluency.
The attrition data behind each stage is more lopsided than most learners expect. The drop-off doesn't spread evenly across the climb; it concentrates at specific points, and those concentrations tell you something important about the kind of problem each one actually is.
Table of Contents
- Name the Obstacle, Not Yourself
- Five Stages & Where People Quit
- A1 → A2
- A2 → B1
- B1 → B2
- B2 → C1
- C1 → C2
- Obstacles by Type
- The Polyglot Advantage
- Your Next Move
- Tools for Each Obstacle
Name the Obstacle, not Yourself
Most learners quit in the fog. Something stops working, progress slows down, and the mind rushes to invent a personal explanation: maybe I'm not built for this, maybe I started too late, maybe I just don't have the discipline. But most stalls are not identity problems. They are navigation problems.
That is what naming the obstacle changes. Each obstacle poses a new challenge, and each one asks for a different response. Once you can see that, the problem becomes much less mysterious.
So the next time the language feels heavy, the question is not what is wrong with me? It is which obstacle is this, and what tools should you use to overcome it? That question is enough to turn frustration back into movement.
The Five Stages of Language Learning, and Where Most People Quit
Note: these figures might look daunting, but they include people who dabble in high school. A self-motivated learner with the right tools and mindset can become fluent at much higher rates.
The CEFR scale runs from A1 to C2. Out of every 100 people who start learning Spanish, here's roughly how many clear each transition:
| Transition | Learners who clear it (of 100 starters) |
|---|---|
| A1 → A2 (elementary) | ~45 |
| A2 → B1 (intermediate) | ~12 |
| B1 → B2 (upper intermediate) | ~4 |
| B2 → C1 (advanced) | ~1 |
| C1 → C2 (mastery) | ~0.2 |
Estimates based on ACTFL/CASLS/FSI proficiency research applied to US learners; see full funnel analysis.
B2 (upper intermediate) is where the language reliably starts to function as a tool rather than an exercise. Most popular usage calls this "fluent." We don't. Real fluency is a stronger claim than B2 supports, and arguably stronger than C1 supports too: CEFR can certify your competencies, but it can't certify that you've made the language home, and a test result is not lived skill. Throughout this site we use "fluency" sparingly, and never below C1.
Only 4 out of 100 reach B2. The question is which obstacles stopped the other 96, and why.
Beginner language learning obstacles (A1 → A2)
Who falls here: Half of all starters.
The A1 → A2 (elementary) transition is where most language learners quietly disappear. Not because the learning is hard. At this stage, it's fast and rewarding. Every lesson is obviously a lesson. You're stacking bricks, and you can see each one.
Finding a reason that survives the requirement — The Interest Obstacle
The problem is whether you actually care about this language. A huge proportion of A1 learners didn't choose to be there: they were dropped in by a school requirement, sitting in a class designed to satisfy a mandate rather than build real competency. Grammar drills. Rote vocabulary. No authentic content. No reason to care. The moment the requirement is met, they leave. Another huge proportion did choose to be there but only out of mild curiosity: a vacation that never materialized, a New Year's resolution, novelty. That kind of curiosity doesn't internalize either.
This isn't really a tools problem. It's an interest and context problem. The research backs this up: in the US, most students exit the foreign language system at somewhere between A1 and A2, having satisfied the minimum required exposure without ever building the internal interest that sustains continued learning. Dörnyei's L2 Motivational Self System gives this a precise diagnosis: learners with only an Ought-to L2 Self (obligation) fall here; learners with an Ideal L2 Self (a vision of themselves using the language) clear it.
The fix for this obstacle is upstream: you need a reason to be here that survives the end of the school year. Travel with stakes, a relationship, a career, the satisfaction of acquiring a skill: any of these works. Learners who clear the Interest Obstacle typically do so because they want the language, not because someone assigned it.
How to choose your first language learning method — The Where-to-Start Obstacle
This obstacle catches the self-starter who's already resolved the desire question. Which app, which textbook, which method, sometimes which language at all. Optimization at A1 is productive-looking procrastination; any reasonable choice beats months of comparison, and switching costs are low.
Building a daily language learning habit — The Habit Obstacle
Learners who are interested still fail to build the daily routine that converts interest into hours. Habit-formation research (Lally et al., 2010) places automaticity at ~66 days of consistent practice; most school-system learners never get there.
Intermediate language learning obstacles (A2 → B1)
Who falls here: About 73% of learners who make it past A2. By total person-hours invested, this is actually the largest attrition event in the entire CEFR funnel (see the attrition table for the math).
The second obstacle is harder to see coming. At A2 (elementary), you've got the basics. You can survive a trip, handle simple transactions, and, in favorable conditions, have a slow conversation. Life intervenes. The class ends. The app gets deleted. And unlike the Interest Obstacle, this one catches learners who actually wanted to learn.
When native speakers switch to English mid-conversation — The Escape Hatch Obstacle
You stumble through three sentences and the native speaker, kindly, switches to English. The conversation continues, but the production push that would have moved you forward is gone. Swain's Comprehensible Output Hypothesis (1985) named this directly: you can't consume your way to fluency. Clearing it takes partners who refuse the off-ramp.
Why motivation collapses after the basics — The Motivation Obstacle
The work has changed shape at A2 → B1. It's no longer basic phrases and basic grammar. It's listening at native speed for an hour and only catching half. It's discovering there are three past tenses, and another tense waiting after them, and then the subjunctive. Visible-progress dopamine slows to a trickle. The interest that got you to A2 isn't enough; you need durable motivation (Dörnyei's Ideal L2 Self doing real work) to keep going.
What's happening structurally is that the American school system has essentially run out of runway. CASLS research found that the majority of students who complete four years of high school language study exit at A1 to A2; the system caps out before B1. Getting to B1 (intermediate) requires somewhere between 300 and 500 hours of quality study for a language like Spanish. Most students accumulate classroom time without the authentic, engaging practice that actually builds proficiency.
The qualitative version of this shows up in Evans and Tragant's 2020 study of 154 adult learners who dropped out versus 106 who kept going. The dropouts consistently named the same three reasons: poor teaching practice, too little speaking practice, and a gap between their desired level and the "stagnant" level they felt they were stuck at. The third reason is the underlying mechanism for this whole transition. When learners can't see themselves moving, they begin to suspect they've hit a personal ceiling, and they quietly leave.
The learners who clear the Motivation Obstacle keep going after the formal environment ends: they find a conversation partner, consume media in the target language, or use a system that keeps them accountable and progressing. The obstacle is real, but it's navigable. You just have to actively choose to keep moving, with a reason that holds up when the dopamine doesn't.
Learning a language without a class or curriculum — The Self-Direction Obstacle
Holec's (1981) work on learner autonomy becomes binding at A2 → B1: with no class and no curriculum setting the next lesson, learners must self-direct (choose materials, build a study plan, find conversation partners) without scaffolding. Most learners weren't trained to do this. The ones who quit at A2 → B1 aren't unmotivated in general; they're motivated people whose motivation was supported by external structure that just disappeared.
Upper intermediate language learning obstacles (B1 → B2)
Who falls here: About 67% of learners who make it to B1.
This is the one that matters most. Not because it catches the most learners in absolute numbers (the earlier obstacles do that), but because it catches the learners who deserve to make it. These are people who committed, put in hundreds of hours, built real skills, and arrived at a level that feels surprisingly close to whatever destination they'd imagined.
Why most language learners stop improving — The Intermediate Plateau
At B1 (intermediate), you can handle a trip. You can follow a conversation if people don't talk too fast. You can order coffee, ask for directions, and, if you're being kind to yourself, feel impressive at dinner parties when you drop into the language for a sentence or two. B1 is real progress.
And then most people stop progressing.
Here's why this obstacle is different from the others, and why it's so much harder to clear.
Why the B1 to B2 jump takes longer than any other stage
Cambridge English's guided-learning-hour estimates put cumulative study at roughly 180–200 hours to reach A2, 350–400 to reach B1, 500–600 to reach B2, 700–800 to reach C1, and 1,000–1,200 to reach C2. The phase durations are roughly similar level to level, but the work changes character at each step. At A1–A2 every hour adds visible vocabulary or grammar. At B1, the same hour produces no externally visible gain. The hours aren't necessarily compounding; the experience of putting them in is.
Why B1 feels like arrival, and that's the trap
There's a seductive quality to B1 that makes this language obstacle uniquely dangerous. You're good enough to enjoy the language. You can watch a show with subtitles and follow it. Travel feels fun rather than anxious. People compliment your Spanish. (Wow, it's amazing. It isn't, but B1 is impressive to people who've never tried.)
This is the intermediate plateau, one of the most well-documented phenomena in SLA research. The academic term is "fossilization": the point where learners' errors stop self-correcting because the language is good enough to communicate, removing the feedback pressure that drove earlier improvement. Your brain is efficient. It stops refining what's working. Patterns solidify, often incorrectly, and learners get stuck producing the same slightly-off grammar indefinitely.
The "plateau" framing is the popular name and we use it here because most learners will recognize it; pedagogically, though, it primes you to wait the terrain out rather than do something specific, and that's why we argue elsewhere for treating it as a stack of named, navigable obstacles instead.
Meanwhile, the gap between B1 and B2 is enormous in ways that don't feel enormous from the inside. A B1 Spanish speaker understands roughly 60–70% of an authentic conversation. That sounds close. But "missing 30–40% of what's being said" is the difference between following a movie and being lost every time someone makes a joke or speaks at speed. It's the difference between "I can work in this language" and "I absolutely cannot work in this language."
The plateau isn't visible from the B1 side. That's what makes it so dangerous.
Same pieces, more arrangements: how intermediate progress actually works
Here's what's actually happening cognitively at this stage, and why none of it looks like progress from the inside. With the same set of Legos, you learn how to assemble them into many different structures, and that is the skill being built at B1 → B2. The work isn't acquiring new pieces. It's the ability to rearrange the words you already have into new contexts, intuitively, until the right one arrives without effort. Practicing the same vocabulary across many different contexts is what moves it into fluent, intuitive use. New pieces aren't important for this process. What you're really attaining is the ability to take the pieces you have and assemble them into diverse structures without thinking about it.
That's why this stage feels invisible. A new word is something you can point to: I learned "sacar" today. A new arrangement, like using sacar across enough contexts that it stops being a word and becomes an instinct (taking out trash, taking a photo, getting cash out of the ATM, getting a grade out of an exam), looks like nothing from the outside. The collocation-acquisition literature describes this directly: the bulk of intermediate-to-advanced gain comes from the same vocabulary getting used in more varied contexts, not from new vocabulary being added. Hours that feel wasted at this stage are actually compounding. You just can't see them yet.
Why the plateau is a systems problem, not a motivation problem
Unlike the first two obstacles, the Intermediate Plateau isn't about motivation or persistence. Learners who reach B1 are committed; they've already proven that. The problem is that the tools they've been using were designed for beginners. Apps, courses, classrooms: they're built for the A1–B1 stretch where the market is largest. They're very good at getting learners to B1. Then they quietly run out of useful things to offer, even while keeping learners busy with content that's no longer moving the needle.
Getting from B1 to B2 requires a different approach: comprehensible input calibrated to where you actually are (not beginner content with a B1 label), high-volume exposure to authentic native-speed speech, forced production without a safety net, and vocabulary development that goes deep rather than wide. None of this comes from a worksheet. And most apps aren't built for it.
For the field guide to this stage in particular, see escaping the intermediate plateau.
Trusting invisible progress in language learning — The Consistency Obstacle
The secondary obstacle underneath the plateau is Consistency: the work is happening, but the gauge feels broken. This is the doubt that kills learners who don't explicitly quit; they slowly stop showing up. The fix isn't more motivation, it's evidence: measurement systems that turn the gauge back on.
Advanced language learning obstacles (B2 → C1)
Who faces this: The ~4 in 100 who've reached B2 (near fluency). The drop is very steep: ~75% of B2-reachers fail to make C1.
If you've reached B2 (upper intermediate), the character of the challenge shifts entirely. You're no longer acquiring the language. You're deepening it.
Going deep on the words you already know — The Depth Obstacle
The question this obstacle asks: can you go deep on the words you already have, instead of accumulating more? Vocabulary breadth has been adequate since B2. What remains is depth: collocations, register tags, multiple senses, the specific words that travel together as formulaic chunks. Wray's research on Formulaic Language (2002) names this directly: native-like fluency is built largely from prefabricated idiomatic chunks, and acquiring them isn't mechanical work. It's long, patient consolidation that doesn't feel like learning.
Catching cultural references and shared context — The Culture Obstacle
You follow every word, but you miss the cultural layer underneath: pop-culture references, regional in-jokes, the historical and political shorthand that natives are actually exchanging when they talk. Linguistic depth alone doesn't get you here. Cultural depth requires its own intake: the books, films, news cycles, public arguments, and shared touchstones that natives absorbed across childhood and adulthood.
Knowing when to use formal vs casual language — The Register Obstacle
When vos lands and when usted is right; when slang is welcome and when a textbook-correct word is socially weird. Hymes's Communicative Competence (1972) names register as a defining advanced-learner skill. The fix is targeted attention to social context, not more volume.
Most B2 learners don't push further because B2 is enough for almost everything they wanted the language for. This is a perfectly reasonable choice. The Depth Obstacle matters for learners with specific professional, academic, or personal goals that require C1-level command, and for those learners, it's real and worth planning for.
Language mastery obstacles (C1 → C2)
First: if you're here, take a moment. C1 is extraordinary. You think in the language. You dream in it. You reach for a word and it's there. For almost every purpose a person could have (work, relationships, literature, travel, daily life), C1 is the summit. Most people who get here are genuinely, permanently changed by it.
Living in your second language — The Deep Immersion Obstacle
C2 is the icing, and the question this obstacle asks is gentler than the ones before it: can you arrange your life so the language stays in it? Structured learning ran out of return at C1. What carries you from C1 to C2 is years of lived experience in the L2 environment: the books you love, the friends you keep, the work you do, possibly the family you build. The SLA literature on ultimate attainment (Birdsong, Ortega) is consistent that this final stretch is predicted almost entirely by the depth and duration of naturalistic exposure. At C1, the good news arrives: the language is fun enough that the exposure doesn't feel like work. You're not grinding anymore. You're just living with it, and getting better by doing so.
Preventing language loss after fluency — The Maintenance Obstacle
At C1 you finally have enough language to lose: skip a few months and the productive vocabulary thins first. The fix is light: a partner you talk with, things you read, shows you watch. At this level it isn't a chore; it's just use.
Obstacles by Type: How the Same Map Looks From a Distance
Zoomed out, the obstacles cluster by type, and each type wants a different kind of fix.
| Obstacle | Transition | Type |
|---|---|---|
| Interest | A1 → A2 | Psychological |
| Where-to-Start | A1 → A2 | Strategic |
| Habit | A1 → A2 | Practice |
| Escape Hatch | A2 → B1 | Practice |
| Motivation | A2 → B1 | Psychological |
| Self-Direction | A2 → B1 | Practice |
| Intermediate Plateau | B1 → B2 | Linguistic |
| Consistency | B1 → B2 | Psychological |
| Depth | B2 → C1 | Linguistic |
| Culture | B2 → C1 | Cultural |
| Register | B2 → C1 | Linguistic |
| Deep Immersion | C1 → C2 | Environmental |
| Maintenance | C1 → C2 | Practice |
The Psychological row reads repetitive but isn't: Interest, Motivation, and Consistency are three different questions in sequence (do I want this?, is my reason strong enough?, can I trust work I can't see?). Practice obstacles want behaviors; linguistic ones want tools. Type matters as much as level.
The Polyglot Advantage: Seeing Every Obstacle Coming
People who've become fluent in a second language learn subsequent languages dramatically faster, and the effect persists even after you control for linguistic overlap. What's happening is that experienced learners have internalized the path: they know the obstacles are coming, roughly where each one sits, and that the Intermediate Plateau is a different kind of problem than the ones before it. They don't panic at B1, and they don't mistake the intermediate plateau for a ceiling. The polyglot's edge isn't talent, it's a map. (We dig into the research on this in are polyglots actually different?.)
Your Next Move
The map's the easy part. Now look back at it. Thirteen named obstacles, each with a different shape and a different fix. The mistake most learners make isn't a moral one; it's diagnostic. They reach for "more discipline" when the obstacle wants a routine, for "more grammar" when it wants production, for "more vocabulary" when it wants depth. The strategies aren't bad. They're just aimed wrong.
Match the Right Tool to Each Language Obstacle
Since every obstacle wants a different fix, the wrong move is a single generic app aimed at none of them. Or worse, one that actually sets up beginners to fail at the intermediate level through over-gamification or shallow design. A tool that actually clears language obstacles has to do a few things a beginner app never will:
- Diagnose which obstacle you're on, instead of serving the next canned lesson.
- Keep input calibrated to your real level, so the Intermediate Plateau doesn't disguise progress as failure.
- Build output in from day one, so the Escape Hatch can't close on you.
- Make the invisible work visible when the gauge feels broken.
For language obstacles, Atlas Runa is the app polyglots built after clearing this exact map in language after language, then turned into a system that picks the right tool at the right moment. It breaks the climb into hundreds of micro-obstacles and matches each to a fix: calibrated comprehensible input that stays inside your comprehension band, output baked in from the start so speaking never gets deferred, a browser extension that carries the same tracking onto the YouTube and articles you'd read anyway, and progress stats that surface the gains the plateau hides. Atlas Runa is the tool that names the obstacle in front of you and hands you the fix that fits it. The method page is the longer version.
Data on learner drop-off rates and CEFR level attainment drawn from ACTFL proficiency research, CASLS studies, and FSI Language Difficulty Rankings. Full funnel analysis: Out of 100 People Who Start Learning a Language, How Many Actually Get Fluent?