Why word lists fail and what works instead

Most language learners start the same way. You find a frequency list or a pre-made Anki deck, and you start grinding through it. "Casa = house. Perro = dog. Comer = to eat." You get through fifty words in a day. A hundred the next week. It feels like progress. You can look at a word and produce the translation. The number on your flashcard app keeps going up.

Then you try to watch a TV show or have a conversation, and those words are nowhere to be found in your brain. You know you studied "comer" last Tuesday. You can picture the flashcard. But when someone says "vamos a comer algo" at normal speed, it doesn't register. The word you "learned" stays trapped behind a wall of decontextualized memorization.

When you learn "correr = to run" from a word list, you're storing a single mapping between two strings. That mapping is fragile. You don't know what prepositions follow "correr" in different situations. You don't know whether it sounds formal or casual. You don't know that "correr un riesgo" means to run a risk, or that "correr la voz" means to spread the word. None of that is on the flashcard.

Norbert Schmitt's research (2008) is pretty clear on this. Learners who study words in isolation develop passive recognition but struggle to use those words productively. They can pick "correr" out of a multiple-choice list. They can't produce it in speech, because it was never connected to any meaningful context. It's floating free, attached to nothing but a translation.

Then there's polysemy. Most useful words have multiple meanings. "Run" in English can mean to jog, to operate a business, to flow, to cost, or to flee. Learning "run = correr" gives you one mapping. The other four meanings are invisible until you hit them in real language — at which point the single mapping you memorized actually gets in the way.

When you learn a word from a video, a podcast, or a story, it arrives with good baggage. You hear how it's pronounced at natural speed. You see which words surround it. You absorb the grammar implicitly — the verb ending tells you tense and person, the word order tells you about syntax, the register tells you whether this is how friends talk or how newsreaders talk. Your brain picks all of this up from the input without you having to think about it explicitly.

Nation (2001) found that words typically need 10 to 12 encounters in context before they move from recognition to productive use. A single flashcard review doesn't count as a contextual encounter. Seeing "correr" in sports commentary, then in a recipe ("dejar correr el agua"), then in a thriller ("tuvo que correr para salvar su vida") does. Each encounter adds another layer to your mental representation of the word.

Stuart Webb (2007) reinforced this: even a few contextual encounters dramatically improved learners' ability to use words in new sentences compared to learners who only studied word-translation pairs. The contextual group didn't just know what words meant. They knew how words behaved.

One contextual encounter is better than a bare word list, but it's not enough on its own. If you only ever see "run" in sports contexts, your mental model of the word will be narrow. You need to see it in a recipe ("let the water run"), in a business podcast ("she runs three companies"), and in a casual conversation ("I ran into Maria yesterday"). Each new context stretches the word's meaning in your head, building a richer and more flexible representation. This is why learning from varied content works so much better than a single textbook, which tends to use the same vocabulary in the same narrow domain. See how comprehensible input drives acquisition for more on why varied exposure matters.

Krashen (1982) made this argument decades ago: acquisition happens when learners are exposed to a large volume of understandable messages in the target language. Not when they memorize rules or lists, but when they process meaning. The more varied the input, the more robust the acquisition. A learner who watches cooking shows, reads news articles, and listens to podcasts builds a vocabulary that transfers across situations. A learner who drills a single frequency deck builds a vocabulary that transfers to the next flashcard session.

Context gives you the initial encoding. It's how words enter your memory with depth and connections. But without review, even well-encoded words fade. That's where spaced repetition comes in. When you review a word using SRS, the key is that the review includes the original context. You don't just see "correr = to run." You see the sentence you found it in, maybe with an audio clip from the video where you first heard it. The context travels with the word, so every review is a mini contextual encounter.

This is the core idea behind sentence mining: save whole sentences, not isolated words. When you review the sentence "tuvo que correr para salvar su vida," you're not just reviewing "correr." You're reviewing the past tense of "tener," the construction "tener que + infinitive," the preposition "para" used for purpose, and the reflexive "salvar su vida." One sentence, five learning opportunities. A word list gives you one. The math is obvious.

Langadoo was built around this principle. You watch YouTube videos with dual subtitles and click words you want to learn. Each word is saved with its full sentence, translation, and audio from the original video. You read AI-generated stories calibrated to your vocabulary, where new words appear in readable context. Every word you save, from any source, enters an SM-2 spaced repetition queue where the sentence context is always present during review. You can also read more about how Langadoo compares to Anki as a vocabulary tool.

The result is that your vocabulary grows the way it does in natural language acquisition: words are always connected to meaning, grammar, and sound. You don't end up with a thousand words you can translate on a flashcard but can't recognize in speech. You end up with words you've heard, read, and reviewed in context, which means you can actually use them.

Start building vocabulary from real videos, stories, and context. Free plan available, no credit card required.

References: Nation, I.S.P. (2001). Learning Vocabulary in Another Language. Cambridge University Press. Schmitt, N. (2008). Review article: Instructed second language vocabulary learning. Language Teaching Research, 12(3), 329-363. Webb, S. (2007). The effects of context on incidental vocabulary learning. Reading in a Foreign Language, 19(2), 232-245. Krashen, S. (1982). Principles and Practice in Second Language Acquisition. Pergamon Press.