What a flashcard can't teach you

We should start by giving flashcards their due, because we're about to argue for keeping them in their lane. The memory science behind spaced repetition is some of the most solid in all of psychology, and tools like Anki deserve their devoted following. We're not here to bury the flashcard. We're here to say what it can and can't do — and why StepText leans on reading instead.

The part that's genuinely true

Two findings anchor everything. The first is the spacing effect: you remember more when your practice is spread out over time than when it's crammed. A 2006 review by Cepeda and colleagues pooled hundreds of experiments and found the effect robust across materials and timescales; a nine-year study of foreign-language vocabulary by Bahrick found wider gaps produced more durable memory per session. The second is the testing effect: retrieving something from memory strengthens it far more than re-reading it. Roediger and Karpicke's 2006 experiments showed restudying wins in the first few minutes and loses badly after a week, and their 2008 study in Science found that, once a word was learned, repeated recall kept it while repeated review did almost nothing.

This is what spaced-repetition software automates: schedule each item so you're asked to recall it right as you're about to forget it. Anki's modern scheduler is fit on hundreds of millions of real reviews. The engineering is excellent and the underlying science is sound. So far, so good for the flashcard.

The part the flashcard quietly skips

The trouble isn't the schedule. It's what fits on the card. Ask an applied linguist what it means to "know a word" and you don't get a definition — you get a framework. Paul Nation lists around nine kinds of knowledge bundled into a single word: its spoken and written form, its meaning (often several), the other words it tends to appear with, its register, its grammatical behaviour, and the difference between recognising it and being able to use it. A front-and-back card drills exactly one of these — the link from written form to a single meaning, in the receptive direction. That's a real and useful slice. It is also a thin one.

Stuart Webb's research makes the point concrete: different aspects of word knowledge grow at different rates, and informative contexts produce deeper, more many-sided knowledge than bare word pairs. You can have a word "mature" in your flashcard deck and still not know which preposition follows it, whether it's formal or slang, or how to deploy it in a sentence of your own. The card said you knew it. The card was measuring the wrong thing.

Reading does the spacing for you

Now the quietly remarkable part. Read regularly and you get the spacing effect for free, with none of the deck maintenance. Useful words are frequent words, so they come back again and again across what you read — naturally distributed over days and weeks, each time in a slightly different context that adds another facet of meaning. This is why "narrow reading" — staying with one author, topic, or series — is such good advice: it recycles the same high-value vocabulary at a natural cadence.

The numbers back this up. A 2023 meta-analysis found that spaced encounters in reading were substantially more effective than massed ones — the spacing effect, showing up automatically inside a reading habit. You're not just learning words while you read; you're rehearsing them on an expanding schedule without ever opening a scheduler.

An aside on the curve every study app loves

You've seen the graphic: Ebbinghaus's "forgetting curve," memory decaying along a tidy line, review heroically flattening it. The decay is real — a faithful 2015 replication reproduced Ebbinghaus's curve remarkably well. But the original was one man memorising nonsense syllables, and the polished graphic hides the interesting bits — like the bump in retention after a night's sleep. The neat percentages and the one-curve-fits-everything story are marketing, not measurement. While we're at it: the folklore that ever-expanding intervals are essential is shakier than people think — evenly spaced review often does just as well for long-term memory. A learning app ought to be willing to debunk learning-app clichés, including its own category's.

Where review genuinely earns its place

This isn't an argument that spaced review is useless — it's an argument about scope. There is a band where deliberate retrieval is exactly right: the stubborn handful of words that context alone keeps failing to fix, the ones you've met ten times and still blank on. A 2021 study by Nakata and Elgort drew the line precisely: spacing helps the explicit, "I can state what this means" kind of knowledge — and it's the implicit, intuitive feel for a word that reading builds. The two are complementary, not rivals.

So that's the division of labour StepText is built around. Reading does the heavy lifting: breadth, depth, and the natural spacing that turns exposure into memory. The review feature is deliberately small — a way to flag the words that aren't sticking and bring them back for a few effortful recalls. It's a rescue tool, not the main course.

If you love Anki, keep it; it's superb at the narrow thing it does, and we're not trying to replace it. But if the deck has started to feel like a second job that somehow isn't translating into reading real sentences, that's the gap this is about. A flashcard can teach you that a word exists. It takes a few hundred sentences to teach you the word.

Sources

1. Cepeda, N. et al. (2006). Distributed practice in verbal recall tasks: a review and quantitative synthesis. Psychological Bulletin.
2. Bahrick, H. et al. (1993). Maintenance of foreign language vocabulary and the spacing effect. Psychological Science.
3. Roediger, H. & Karpicke, J. (2006). Test-enhanced learning. Psychological Science.
4. Karpicke, J. & Roediger, H. (2008). The critical importance of retrieval for learning. Science.
5. Nation, I.S.P. What it means to know a word.
6. Webb, S. (2007). The effects of repetition on vocabulary knowledge. Applied Linguistics.
7. Webb, S. (2008). The effects of context on incidental vocabulary learning. Reading in a Foreign Language.
8. Webb, Uchihara & Yanagisawa (2023). Incidental vocabulary learning: a meta-analysis. Language Teaching.
9. Murre, J. & Dros, J. (2015). Replication and analysis of Ebbinghaus' forgetting curve. PLoS ONE.
10. Karpicke, J. & Roediger, H. (2007). Expanding vs. equally spaced retrieval. JEP:LMC.
11. Nakata, T. & Elgort, I. (2021). Effects of spacing on contextual vocabulary learning. Second Language Research.