Ask this question in any language-learning forum and you will start a fight. One camp tells you to delete Anki, pick up a stack of books, and let your vocabulary grow the way a child’s does, by meeting words over and over in real sentences until they stick. The other camp tells you that reading without review is a sieve, that you will meet a word, feel that warm flicker of recognition, and forget it three pages later, and that the only honest fix is to put it on a card and review it on schedule.
Both camps have a point, and both tend to argue against a cartoon version of the other side. The “just read” people imagine flashcard users as joyless robots drilling decontextualized word lists for hours. The “review or forget” people imagine readers drowning in books they barely understand, fooling themselves that exposure alone is teaching them anything. Neither caricature is fair, and neither answers the real question, which is narrower and more useful: if you already read a lot in your target language, does adding a spaced repetition habit pay for the time it costs?
That is a question about trade-offs, not ideology. Spaced repetition is real, it is one of the best-studied findings in the science of memory, and it does something reading does not do reliably. Reading is also real, and it does several things spaced repetition cannot do at all. The interesting part is where they overlap, where they leave gaps, and how to combine them so you get the durable recall of flashcards without letting card-making eat the reading time that made you want to learn the language in the first place.
This article tries to give the fair version of the answer. We will look at what spaced repetition genuinely does, what reading does that no flashcard system can, where each approach leaves holes, who should keep their deck and who can close it, and a practical middle path for everyone in between. No deleting Anki out of spite, no defending it out of habit.
What Spaced Repetition Genuinely Does
Start with the part that is not up for debate. Spaced repetition works, and we have known why for a long time.
In 1885 the German psychologist Hermann Ebbinghaus published the first experimental study of memory, where he memorized lists of nonsense syllables and tracked how fast he forgot them. He produced what we now call the forgetting curve: memory of new material drops off quickly at first, then more slowly, in a roughly exponential decay. He also noticed something more useful. When study sessions were spread out over time instead of crammed into one block, the same amount of effort produced far better long-term retention. That finding, the spacing effect, has held up across nearly 140 years of replication on free recall, recognition, cued recall, and more. Later researchers, including Robert Bjork in the 1960s and 1970s, formalized it into the modern theory of spaced practice.
Spaced repetition software turns that principle into a schedule. Anki, the best-known tool, is free and open source on desktop and Android, and it shows you each card just as you are about to forget it. Get it right and the interval before the next review grows; get it wrong and the interval shrinks. Historically Anki used a variant of the SM-2 algorithm, designed for SuperMemo in the late 1980s, and in 2023 it added the option of a newer scheduler called FSRS, tuned on hundreds of millions of real reviews to hit a target retention rate with fewer cards shown. (One detail worth knowing before you start: the iPhone and iPad app, AnkiMobile, is a paid one-time purchase, while the desktop apps and AnkiDroid on Android are free.)
What all of this buys you is deliberate, durable recall of specific items you choose. If there are eighty words you have decided you must know cold, a tax vocabulary for an exam, the irregular verbs that keep tripping you, the medical terms for your job, spaced repetition is close to the most efficient way ever invented to drive them into long-term memory and keep them there. It also gives you control. You decide what goes in the deck, so nothing important falls through the cracks because you happened not to meet it often enough.
In Paul Nation’s well-known framework of the four strands of a balanced language course, this is the language-focused learning strand: the deliberate study of vocabulary, grammar, and pronunciation. Nation argues a good course gives roughly equal time to four strands: meaning-focused input (reading and listening for content), meaning-focused output (speaking and writing to communicate), language-focused learning (the deliberate study), and fluency development (getting faster with what you already know). Spaced repetition is a tool for exactly one of those four, and it is very good at its job. The mistake is expecting it to cover the other three.
What Reading Does That Flashcards Cannot
Now the other side, and it is just as real.
Reading delivers volume that no deck can match. An hour of comfortable reading puts thousands of words in front of you, most of them familiar, a handful new, all of them in living sentences. A heavy Anki session might rehearse a hundred cards. The numbers are not close. Vocabulary growth at the size most learners need, several thousand words to read freely, runs on exposure counts that only reading and listening can supply. We wrote about the raw scale of this in how I learned 3,000 words through reading, and the arithmetic is hard to argue with.
Reading also teaches the things a single-word card strips out. It teaches collocation, which words actually go together, so you learn that you make a decision but take a photo, not the reverse. It teaches register, whether a word is street slang or courtroom formal. It teaches the dozen shades of a word that no two-sided card can hold, because you meet it in twenty different contexts and your brain quietly averages them into real understanding. A flashcard says a word means one thing. A book shows you what it does.
And reading builds automaticity, the ability to understand without stopping to translate. That only comes from meeting the same words again and again in fast, meaningful reading, which is Nation’s fluency development strand. A card trains you to retrieve a word when prompted in isolation. Reading trains you to recognize it instantly as it flies past inside a sentence you actually care about, which is a different and arguably more important skill. We make the broader case in why reading is the best way to learn a language, and the short version is that reading is the only activity that scales meaning-focused input the way fluency requires.
There is one more thing reading does that decks structurally cannot: breadth. You cannot put a word on a card until you have met it somewhere. Reading is the meeting place. It constantly hands you new candidates, including the rare and oddly specific words that no premade frequency deck would ever include but that happen to matter for the book in front of you. Decks are downstream of input. Reading is the input.
The Real Tension: Incidental Versus Deliberate
So if reading supplies the volume and the context, and spaced repetition supplies the durable recall, why not just read and trust that repetition inside books will do the same job as a deck? This is the heart of the disagreement, and it deserves an honest answer rather than a slogan.
Learning a word from reading is incidental learning. You were reading for the story, the word came along for the ride, and over enough encounters it stuck. The research broadly supports that this works, but it works slowly and unevenly. Estimates vary, but it commonly takes somewhere on the order of a dozen or more meaningful encounters with a word before it is reliably yours, and high-frequency words get there fast while mid- and low-frequency words may take a very long time, because by definition you meet them rarely. We dug into the encounter math in how to remember the words you read, and the uncomfortable conclusion is that pure reading retains the common words well and lets a lot of the less common ones slip.
Spaced repetition is deliberate learning. You decided this word matters, so you schedule encounters with it instead of waiting for chance to supply them. That fixes the exact weakness of incidental learning: the word you need but rarely meet. It also fixes recall under pressure, because a card forces active retrieval, while reading mostly trains recognition. Recognizing a word when you see it is easier than producing it when you need it, and that gap is where a lot of readers get stuck.
Which is why the honest summary is that each approach leaves a real gap. Reading alone, even a lot of it, tends to leave you with strong passive recognition of common vocabulary and weak active command of anything rarer, especially for production. You will understand far more than you can say. Spaced repetition alone leaves you with a set of words you can recite on cue but cannot use fluently, because you never met them at speed in real context, plus it does nothing for reading fluency itself and tends to feel like a chore long before it has taught you enough words to read freely. The question is never which one is correct. It is which gap matters most for you right now.
Who Should Keep Spaced Repetition, and Who Can Drop It
The fair answer depends heavily on where you are and what you want, so here is the honest sort.
Keep spaced repetition if you are a beginner. Early on you do not yet have enough vocabulary to read comfortably, and reading only teaches efficiently once you understand most of what is on the page. A deck of the first thousand or two high-frequency words gets you to that threshold faster than struggling through text where every other word is a mystery. Decks are a ladder to the level where reading takes over. They are a poor permanent home.
Keep it if you are studying for an exam with a fixed word list or deadline. When you must know a specific, defined body of vocabulary by a specific date, deliberate scheduling beats hoping the right words show up in your books in time. This is the case where the control that decks give you is worth the most.
Keep it if your goal is production, speaking and writing, not just comprehension. Reading is fabulous for understanding and weak for retrieval. If you need to pull words out of your own head on demand, the active retrieval that flashcards force is hard to replace, particularly for words you do not meet often enough in input to overlearn naturally.
You can usually drop it, or shrink it to almost nothing, if you are past the beginner stage, you read a lot in a comfortable difficulty band, and your goal is to read and understand rather than to perform on a test or speak under pressure. At that point reading is already supplying the volume, the context, and most of the repetition. A daily deck of a few hundred cards may simply be redoing, less pleasantly, what your reading does for free. Many advanced learners quietly stop their decks around here and lose nothing, because their books have become their spaced repetition system. If you tend to plateau on apps that drill instead of immerse, the same logic that makes Duolingo frustrating for advanced learners applies to a giant flashcard deck: past a point, isolated drilling stops moving the needle and real input does.
The middle, where most people actually live, is not “deck or no deck.” It is a small, sharp deck fed by reading. That is the practical path.
The Practical Middle Path: Read First, Mine Sparingly
Here is a workflow that respects both the science and your limited time. The order matters.
Read first. Reading is the engine, the source of words, context, and the comprehensible input that does most of the heavy lifting. It is also the part you enjoy, which is what keeps you doing this for years instead of weeks. Protect it. If your study time is an hour, the great majority of it should be reading, not card management. Pick books at a level where you understand most of the words without help, the comfortable band described in how to read without a dictionary, so that reading itself stays mostly fluent and only a thin stream of new words needs capturing.
Mine sparingly. As you read, you will meet new words. Do not card all of them. Carding every unknown word is the single fastest way to build a deck you will come to dread. Instead, capture only the high-value ones: words that appear repeatedly, words that block your understanding of a sentence you care about, words that are clearly common in this kind of writing. A useful rule of thumb is that if a word stopped you twice, it has earned a card. If you have met it once and the sentence still made sense, let it go. Reading will bring it back if it matters, and if reading never brings it back, it did not matter.
Review only high-value items. The deck exists to catch the words that reading alone would lose, the rarer ones, the ones you need to produce, the ones that keep slipping. It is not a transcript of everything you have ever looked up. A lean deck of genuinely useful words, reviewed daily for a few minutes, gives you nearly all the retention benefit of a giant deck with a fraction of the misery and almost none of the risk of quitting. Quality of selection beats quantity of cards every time.
And do not let card-making eat reading time. This is the rule that holds the whole thing together. The moment your review-and-card-creation workload starts crowding out the reading, you have inverted the priorities. The deck is the supplement. The reading is the meal. If the supplement is taking longer than the meal, cut the supplement, not the meal. For a sense of how much reading you actually need at the volume end, how many words to read in a foreign language lays out the targets, and they are large enough that you cannot afford to spend half your time making flashcards instead.
This is also, not coincidentally, close to a balanced version of Nation’s four strands at the individual level. The reading covers meaning-focused input and, as it speeds up, fluency development. The small deck covers language-focused learning. The only strand left is output, which neither reading nor a deck supplies, and which is why people who do both still need to actually speak and write.
Sentence Cards Versus Word Cards
If you do keep a deck, one design choice matters more than most: what goes on the front of the card.
A word card has the bare word on one side and the meaning on the other. It is fast to make and fast to review, and it is exactly the thing the “flashcards are decontextualized” critics complain about, because it strips the word out of every sentence that gave it meaning. Word cards are fine for a small set of high-frequency basics where the word really does map to one clear meaning, and they are efficient when you need raw coverage in a hurry, like a beginner deck.
A sentence card puts the word inside the actual sentence where you met it, usually with the target word highlighted or blanked, and tests you on the word in that context. This is closer to how reading teaches and closer to how you will actually use the word. Sentence cards preserve collocation and register, they remind you what the word does and not just what it allegedly means, and because you mined the sentence from something you were genuinely reading, the card carries a hook of meaning and even memory of the story. They take a little longer to make and review, and that is the trade-off.
For words mined from reading, sentence cards are usually the better choice, because the whole advantage of learning from books, context, is exactly what a sentence card keeps and a word card throws away. The pragmatic compromise many learners settle on: word cards for a small starter set of high-frequency vocabulary when you are racing to the reading threshold, then sentence cards mined from real reading once you are past it. This bridges the distinction the “Anki is dry” critics raise. A sentence card mined from a novel you love is not dry. It is a small piece of that novel, scheduled.
The Burnout Failure Mode
No honest discussion of Anki is complete without the way it most often fails, which is not that it does not work. It is that people quit.
Anki burnout is real and it has a predictable shape. You start strong, you add cards faster than you review them, the daily count climbs, and one morning you open the app to four hundred due cards and feel a small dread instead of motivation. You skip a day. Now it is six hundred. You skip a week. Now the backlog is a wall, the reviews are stale, and the whole thing feels like debt collection. You stop. And here is the quiet tragedy: when people burn out on Anki, they often blame spaced repetition itself and swear off all structured review, when the real culprit was a deck that grew faster than they could sustain.
The causes are mostly self-inflicted and all avoidable. Adding too many cards, especially carding every unknown word from your reading, guarantees a flood. Making cards that are too hard or ambiguous turns each review into a struggle. Setting retention targets too high makes intervals short and reviews relentless. Treating a missed day as failure rather than a normal blip turns a small backlog into a reason to abandon ship.
The defenses follow directly. Cap how many new cards you add per day, well below what feels generous, because today’s new cards are next month’s review load. Mine sparingly, as above, so the deck stays a curated set and not a dumping ground. Keep cards clear and answerable. Accept that a backlog after a break is normal, and just grind it down a bit at a time instead of declaring bankruptcy. Above all, remember that the deck is in service of the reading, so if the deck is making you miserable enough to quit the whole project, the deck is too big and the fix is to cut it, not to cut the language. A tiny deck you actually keep doing beats a huge deck you abandon in three weeks. This is the same lesson that separates extensive from intensive reading: the sustainable, lighter habit you maintain for years beats the intense one you burn out on.
Mining Words From Reading With Lingo7
The hardest part of the read-first, mine-sparingly approach is usually the friction. You are reading on one device, looking words up in a dictionary, copying them into Anki, building cards, and finding sentence context, and every one of those hops is a chance to lose the flow of reading or to give up on the deck. Lingo7 is built to collapse that friction so the workflow happens in one place.
You read real books in the app, and when you meet an unknown word you tap it for an instant translation in context. Words you save from your reading go into a built-in spaced repetition system, so the mining and the reviewing both live inside the reading app. You do not switch tools, you do not copy sentences by hand, and you do not build cards from scratch. The word arrives already attached to the sentence you met it in, which is exactly the sentence-card advantage described above, captured automatically from what you were actually reading.
Synchronized audio lets you hear the text as you read, which adds the listening strand and reinforces the same words through a second channel. Lingo7 supports over 90 languages, runs on iOS and Android, and is free to start, so you can test whether reading plus light, built-in review works for you before committing anything. If you want a specific starting point, the per-language guides like learning Spanish walk through how to begin reading in your target language from your current level. The goal is not to replace deliberate study with magic. It is to make the mine-sparingly half of the workflow cheap enough that you will actually do it, instead of letting good words slip away because making cards was too much trouble.
The Bottom Line
Do you still need Anki if you read a lot? The fair answer is: probably not as much as you think, and rarely zero.
Spaced repetition is one of the most solid findings in the science of memory, and it does one thing better than anything else, driving specific, chosen words into durable recall, which is precisely the thing reading does least reliably for rare and production vocabulary. Reading does the things a deck structurally cannot: volume, context, collocation, breadth, and the automaticity that only comes from meeting words at speed in sentences you care about. Pure reading leaves gaps in active recall and low-frequency words. Pure flashcarding leaves you with words you can recite but not use, and a habit that tends to collapse before it has taught you enough.
So keep a deck if you are a beginner climbing to the reading threshold, if you have an exam with a fixed list, or if your goal is production. Shrink it toward nothing if you are past the beginner stage, read a lot, and mostly want to understand. For everyone in the middle, the answer is not deck versus books. It is a small, sharp deck fed by your reading: read first, mine only the high-value words, prefer sentence cards over word cards, keep the deck small enough that you never burn out, and never let card-making crowd out the reading that made the words worth learning. Anki is a fine tool. Reading is the engine. Use the tool to patch the gaps the engine leaves, and not one card more.