Every learner who decides to read in a foreign language runs into the same wall almost immediately. You open a real novel, the kind a native speaker would read, and within two sentences you are drowning. You look up one word, then the next, then the next, and by the time you reach the bottom of the page you have forgotten how the page started. The book that was supposed to teach you the language has become a chore you dread.
So you do the sensible thing. You buy a graded reader instead, a book written specifically for learners, with controlled vocabulary and simplified sentences. And it works. You can actually read it. You finish a chapter without exhausting yourself. But then a different feeling creeps in. The story is thin. The prose is flat. You can feel that someone has sanded off all the interesting edges to keep you safe, and after a few of these you start to wonder whether you are really learning the language or just learning a watered-down version of it.
This is the real dilemma, and it has nothing to do with willpower. It is a structural trade-off. On one side you have material you can read but do not want to. On the other you have material you want to read but cannot. Boring but easy, or exciting but impossible. Most advice online pretends one side is simply correct, but the honest answer depends on where you are right now and where you are trying to go.
This article lays out what graded readers and original books actually offer, the simple coverage math that decides which one is readable for you today, and a level-by-level framework for choosing between them. It also covers a third option that quietly changes the whole equation: parallel text, which lets you read real books much earlier than the coverage math would normally allow.
What Graded Readers Are
A graded reader is a book deliberately written or rewritten to stay within a fixed vocabulary and grammar budget. The central concept is the headword: a base word the reader is expected to know. A graded reader at a given level promises to use only headwords up to a certain count, plus a small number of words it explains or makes obvious from context. This is what lets a beginner read a whole book without looking up every line.
The major English series make the system concrete. The Oxford Bookworms Library runs across seven levels, from Starter at 250 headwords up through Stage 6 at 2,500 headwords. The intermediate steps are 400, 700, 1,000, 1,400, and 1,800 headwords. Penguin Readers covers a similar span, from a Pre-A1 starter level of around 350 headwords up to roughly 2,500 headwords at B2. The actual book lengths grow with the level. An Oxford Starter title might run around 1,000 to 1,500 words total, a Stage 1 book averages roughly 5,200 words, and a Stage 6 title can reach about 30,000 words, which is a genuine short novel.
The vocabulary ceiling is the whole point. At Stage 1, a learner who knows the 400 most frequent words can read the entire book with very few surprises. The grammar is controlled too: early levels stick to simple tenses and short sentences, and later levels gradually introduce the passive voice, conditionals, reported speech, and longer subordinate clauses. The difficulty is engineered to rise in steps you can actually climb.
Series like these exist for many languages, not just English. Olly Richards’ Short Stories collection, published through Teach Yourself, offers graded short-story books mapped to roughly A2 to B1 on the CEFR, available in Spanish, French, German, Italian, Russian, and others. The Extensive Reading Foundation maintains a scale that lets you compare levels across different publishers, because a “Level 3” from one company is not always a “Level 3” from another. The headword system is the common thread.
It helps to be precise about what a graded reader is not. It is not a textbook, and it is not a dumbed-down children’s book. The best ones are written by real authors who treat the constraint as a craft. But the constraint is always there, sitting between you and the language a native speaker would actually meet.
The Case for Graded Readers
The strongest argument for graded readers is coverage, and coverage is the single most important variable in whether reading helps you at all. When a book is written to your level, you understand the overwhelming majority of the words on every page. That means you can read for meaning instead of decoding, you can read fast enough to enjoy a story, and you can read a lot. Volume is where reading gains come from, and graded readers make volume possible early, when an original book would stop you cold.
This connects directly to the idea of extensive reading: reading large amounts of easy material for pleasure and fluency rather than grinding through hard material word by word. Graded readers are practically built for the extensive approach. Because the vocabulary stays within reach, you meet the same high-frequency words again and again across many books, which is exactly how those words move from “I sort of recognize this” to “I know this instantly.” Repetition in context, at speed, across volume, is the engine.
There is also a confidence dividend that is easy to underrate. The first time you finish an entire book in a foreign language, something shifts. You stop seeing the language as an impossible mountain and start seeing it as a thing you can do. For a beginner, the psychological value of completing a real, bound book, even a simple one, is worth more than the handful of words it teaches. A learner who feels capable keeps going. A learner who feels crushed quits, and the best material in the world cannot help someone who has stopped reading.
Graded readers also remove a hidden tax. With an original book, a huge share of your mental energy goes to looking things up, losing the thread, and rereading. With a level-appropriate graded reader, almost all of that energy goes to the story and the language itself. The friction is low by design, and low friction is what lets a habit survive past the first week. If you are choosing your very first book in a foreign language, a graded reader at or slightly below your level is the safest possible bet.
The Limits of Graded Readers
For everything they do well, graded readers have three real weaknesses, and pretending otherwise sets learners up for disappointment.
The first is the prose itself. Controlled vocabulary means controlled expression. When a writer cannot use the precise word, the surprising metaphor, or the long winding sentence, the language flattens. Subtlety, voice, and style are the first casualties of simplification. Many graded readers read like competent summaries of stories rather than the stories themselves. For an early learner this does not matter, because you are reading for the words, not the art. But as your level rises, the flatness starts to feel like a cage, and reading begins to feel like eating food with no salt.
The second limit is supply. There are thousands of graded readers in English, fewer in major languages like Spanish, French, and German, and very few in everything else. If you are learning a language outside the big commercial market, the well runs dry fast, especially above the lower-intermediate levels. You can burn through the available titles at your level and find that the next level up barely exists. Original books, by contrast, are effectively infinite. Every novel, article, and forum post ever written in the language is fair game, with no publisher deciding what you are allowed to read.
The third limit is the one nobody mentions when you start: graded readers have a ceiling, and you must eventually leave them. The whole system tops out around B2, roughly 2,500 headwords. Real fluency lives well beyond that. At some point the training wheels stop helping and start holding you back, because the only way to learn the full, uncontrolled language is to read the full, uncontrolled language. Graded readers are a bridge, not a destination. A learner who stays on them too long plateaus, comfortable and stuck, reading the same narrow band of simplified prose forever. The point of the bridge is to cross it.
What Original Books Give, and What They Cost
An original book, also called an authentic text, is one written for native speakers with no thought of learners at all. That is precisely its value and its danger.
The value is that it is the real language. Full vocabulary, natural grammar, idioms, register shifts, cultural references, the way people actually write. Nothing has been removed. When you read an authentic book, you are reading the same words a native reader reads, and every word you learn is a word that genuinely appears in the wild. There is no ceiling, because the supply is the entire written output of the language and the difficulty goes as high as the language itself goes.
The motivation is real too. People learn languages to read specific things: a novelist they love, the news, a field they care about, the lyrics of songs that mean something to them. A graded adaptation of a great book is not the great book. Reading the actual text you came for is its own fuel, and motivation sustains the thousands of hours that fluency requires. Some books simply land differently in the original language than any translation or adaptation can manage, and chasing that experience is a perfectly good reason to push into authentic material.
The cost is coverage, and early on the cost is brutal. A native-level novel routinely uses tens of thousands of distinct words, and a beginner knows a tiny fraction of them. Open a real novel at A2 and you might understand half the words on the page, which means you understand essentially none of the meaning. Reading becomes a dictionary marathon: stop, look up, lose the thread, reread, stop again. It is slow, exhausting, and demoralizing, and most learners who try to jump straight into authentic books too early simply give up and conclude they are bad at languages. They are not. They picked material with coverage so low that comprehension was mathematically impossible. To understand why, look at the numbers.
The Coverage Math That Decides Everything
There is a well-known finding in vocabulary research that cuts through the whole debate. Hu and Nation (2000) studied how much of a text a reader needs to know to understand it without help. Their conclusion: you need to know around 98% of the words in a text for comfortable, unassisted comprehension. At 98% coverage, only about one word in fifty is unknown, which is roughly one new word per two or three lines, a rate at which you can usually guess meanings from context and keep reading. The study is not the last word (it rested on a fairly small sample, and later replications have produced mixed results), but the broad picture has held up well: comprehension falls off a cliff as unknown-word density rises. Other researchers put a useful floor around 95% coverage for getting the gist with effort, with 98% as the threshold for reading that feels smooth.
Now apply that to the choice in front of you. A graded reader at your level is engineered to put you at or above 98% coverage. That is literally what the headword system buys. An original novel, for a learner below roughly B2, often sits at 80% to 90% coverage, far under the threshold, which is exactly why it feels impossible. The 98% rule explains the whole experience in one stroke: the graded reader is readable because the math works, and the novel is unreadable because the math does not.
This also reframes the goal. The real question is never “graded reader or original book” in the abstract. It is “what gives me 95% or more coverage today, and how do I keep raising the coverage I get from harder material over time.” Vocabulary size is the lever. The more words you know, the more original books cross into readable territory on their own. If you want to see roughly how many words different thresholds require, we cover that in detail in how many words you need to read in a foreign language. For now, hold onto the simple version: coverage decides readability, and coverage is what you are trying to grow.
Parallel Text: The Bridge Between Them
Here is where the binary choice quietly breaks down. The coverage math assumes you read a text the normal way, with nothing but a dictionary to help. But there is a way to read original books that effectively raises your coverage without changing your vocabulary at all.
Parallel text places the original and a translation side by side, or makes the translation available the instant you need it. When you hit an unknown word in an authentic sentence, you do not stop, lose the thread, and dig through a dictionary. You glance at the meaning and keep moving. Functionally, this lifts your usable coverage. A novel that gives you 88% coverage on your own, far below the comprehension threshold, becomes readable when the missing 12% is one glance away. You are still reading the real language, full vocabulary and natural grammar intact, but the friction that normally makes authentic books impossible is gone.
That is what makes parallel text a genuine bridge rather than a crutch. It does not water down the original; it removes the lookup tax that the coverage math punishes you for. A learner who would otherwise be stuck on graded readers for another year can start reading the books they actually want, at a level where the unaided numbers say they should not be able to. The training wheels are on the bike, but the bike is real.
Used well, parallel text shortens the very gap this article is about. Instead of grinding through graded readers until your vocabulary alone crosses the 98% line on real novels, which can take a long time, you read real novels now and let the volume push your vocabulary up while you go. Used badly, it can become a way to read translations and barely look at the original, so it is not automatic magic. We go through exactly when it helps and when it holds you back in our honest guide to parallel reading. The short version: lean on the translation as a safety net, not as the main event, and your eyes should live in the target-language column.
A Decision Framework by Level
Strip away the theory and the choice comes down to your level and your honest tolerance for friction. Here is a practical map.
Absolute beginner (A0 to A1). Graded readers, full stop. Your vocabulary is too small for the coverage math to work on anything authentic, and an original book will only teach you that reading is misery. Start at the lowest graded level you can find, even one that feels embarrassingly simple. The goal is to finish books, build the habit, and meet the highest-frequency words enough times that they stick. Parallel text can supplement here, but keep the texts very short and very easy.
Lower intermediate (A2 to B1). This is the transition zone, and the smart play is to read on both tracks. Keep reading graded readers at the upper levels for fluency and confidence, but start dipping into easy authentic material with parallel text as your bridge. Short stories, children’s and young-adult books, and simpler contemporary novels are the natural first targets. You will not have 98% coverage on real books yet, which is exactly why the parallel-text support matters at this stage.
Upper intermediate (B2). The balance tips toward originals. You are near the top of what graded readers offer, so they have less and less to teach you. Authentic books with a translation on hand are now your main diet, and you will find that for easier or more familiar texts you can increasingly drop the translation and read straight through. This is the stage where most learners should be actively weaning off graded readers, because staying on them now risks a comfortable plateau.
Advanced (C1 and up). Original books only, read normally. Graded readers have nothing left to give you, and your coverage on most authentic texts is high enough that you read for pleasure and polish, looking up only the rare unknown word, the same way a native reader occasionally does. The bridge is behind you. You crossed it.
Two cautions on the framework. First, level is not a single number across all topics: you might be B2 in everyday narrative and A2 in legal or scientific prose, so match the material to the specific domain, not just your overall CEFR badge. Second, these bands are guides, not gates. The real test is always coverage and friction on the actual book in your hands. If a book feels smooth and you are learning, it is the right book, whatever label it carries. For concrete title suggestions tuned to each stage, see our guide to the best books by language level.
How to Make the Transition
The move from graded readers to originals is where most learners either break through or stall, so it deserves a real plan rather than a leap of faith.
Do not switch cold. The worst version of this transition is finishing your last graded reader on Monday and opening Dostoevsky on Tuesday. The coverage gap is too wide and you will bounce off. Instead, overlap the two. While you are still comfortable in upper-level graded readers, begin a single easy authentic book on the side, read with parallel text, and let the two coexist for weeks or months. The graded readers protect your confidence while the authentic book stretches you.
Choose your first originals for coverage, not prestige. The instinct is to reach for the famous, difficult classic you have always wanted to read. Resist it. Your first authentic books should be the ones with the friendliest language: contemporary fiction over nineteenth-century literature, plot-driven genre novels over dense literary prose, books you have already read in translation so the story carries you past the hard parts. Familiarity is a coverage multiplier, because knowing what happens next lets you guess words you do not technically know.
Reread on purpose. A book you have already finished, returned to a few weeks later, is far easier the second time, and the second read is where a lot of vocabulary consolidates. This is true for both graded readers and originals. Rereading is not a step backward; it is one of the most efficient things you can do.
Track the friction honestly. The right book sits at the edge of comfort: challenging but not crushing, with new words frequent enough to learn from but rare enough that you can still follow the story. If you are looking up more than a few words per page even with help, the book is too hard, so drop down a notch. If you almost never meet a new word, it is too easy to grow from, so step up. You are constantly steering toward that 95% to 98% sweet spot, and the steering never really stops.
Above all, keep reading volume high through the whole transition. The reason reading works at all is covered in why reading is one of the best ways to learn a language: sustained exposure to words in real context, in quantity, is what builds durable vocabulary and intuition. Graded readers, parallel text, and originals are just different tools for keeping that exposure high at different stages. The tool matters less than the volume.
Reading With Lingo7
Most of the practical problems in this article come down to friction and coverage, and that is the gap Lingo7 is built to close. It puts the original text and a translation together, so when you read an authentic book below your unaided coverage level, an unknown word is one tap away instead of a trip to the dictionary. That tap-for-translation is what effectively raises your usable coverage and lets you start real books earlier than the 98% math would normally allow, which is the whole bridge this article describes.
A few things make the bridge sturdier. Audio is synchronized with the text, so you can read and listen at once and train your ear alongside your eye. Words you tap can be saved and sent to spaced-repetition review, so the vocabulary you meet while reading gets reinforced on a schedule instead of fading by the next page. It supports 90+ languages, which matters most for learners outside the big commercial market, where graded readers run out fast and authentic books are the only deep supply you have. It runs on iOS and Android, and it is free to start, so you can test whether parallel reading actually fits how you learn before committing anything.
Lingo7 will not do the reading for you, and no app turns an A1 beginner into a novel reader overnight. What it does is remove the specific friction that makes original books feel impossible, so you can spend your energy on the language instead of on lookups. If your target language has a learn page, it is a good place to see what reading in that language looks like before you start.
The Bottom Line
Graded readers and original books are not rivals. They are two stages of the same path. Graded readers exist because the coverage math is real: below a certain vocabulary, authentic text is genuinely unreadable, and trying to force it just teaches you to quit. They give beginners the one thing nothing else can, a way to read whole books, build volume, and gain the confidence to continue. That is not a compromise. It is the correct tool for the early road.
But graded readers top out around B2, and real fluency lives past that line. The flat prose, the thin supply, and the hard ceiling are all signs that, eventually, you have to leave. Original books are where the language actually lives: full vocabulary, real voice, infinite supply, and the books you actually want to read. The only catch is the coverage gap that makes them brutal early on.
Parallel text is what collapses that gap. By putting the meaning one glance away, it raises your usable coverage and lets you read real books years before raw vocabulary alone would permit, without watering the language down. So the honest answer to “graded readers or original books” is: graded readers first, originals as soon as you can reach them, and parallel text as the bridge that lets you reach them sooner. Start where the coverage math says you can read, keep the volume high, and keep steering toward harder, more real material as your words grow. The boring-but-easy versus exciting-but-impossible trap is not permanent. It is just the part of the journey you read your way out of.