You're reading a message from a Spanish-speaking friend, a line in a graded story, a caption under a photo. A sentence like "se puso a llover a cántaros" doesn't land as meaning first. It lands as a decoding problem: rearrange the words, guess at the idiom, land on "it started raining buckets," then go back and reread the original now that you know what it says. By the time you understand the sentence, you've processed it twice, once in Spanish, once in the English you built to get there. That extra pass is why reading feels so much slower in Spanish than in English, and why it rarely speeds up just from reading more of the same way.
It also explains a specific kind of frustration: finishing a graded reader, or a few seasons of a show with subtitles, and feeling like nothing changed. The exposure was real, but if every sentence got translated on the way in, the brain practiced translation, not reading. More hours of that practice make the translating faster, not unnecessary.
The fix usually gets blamed on vocabulary. More words, the thinking goes, and the translating will stop on its own. It doesn't work that way. People with large vocabularies translate in their heads for years if nothing changes how they practice reading. The habit isn't about how many words you know. It's about which path your brain takes to get from a Spanish sentence to its meaning.
Why translating breaks reading
Fluent readers, in any language, don't decode text word by word. They recognize whole chunks, common phrases, verb forms, set expressions, and pull meaning from them directly, the way you read "I'm starving" without parsing each word on its own. Translating adds a step in between. Instead of Spanish word to Spanish meaning, the path becomes Spanish word to English word to English meaning, and the detour runs through a language the sentence was never written in.
That detour explains two things at once: why reading in Spanish tires you out after a page when reading in English doesn't, and why the tiredness has nothing to do with how hard you're trying. Working memory in a second language already does more than it does in your first, holding unfamiliar grammar and new words in place while it reads. Running every sentence through a second conversion on top of that is what wears a reader out, not a lack of effort or a weak vocabulary.
What reading without translating actually requires
Paul Nation, a linguist at Victoria University of Wellington, found that readers need to already recognize around 98% of the words on a page before they can understand a text without outside help (Nation, 2006). Below that threshold, a reader isn't really reading anymore. They're solving a puzzle: stopping, guessing, checking a dictionary, translating a phrase to confirm a guess. Above it, there's enough familiar material for the brain to infer the rest from context, the same way an unfamiliar English word gets inferred from the sentence around it instead of looked up.
That threshold is also why a novel written for native speakers, picked up at an intermediate level, usually trains the opposite of what it's meant to. Without 98% of the words already known, a reader isn't reading directly, they're translating constantly and calling it reading practice. The habit gets reinforced, not broken, by material that's too hard for where they actually are.
Why gap-fill exercises don't retrain this
Apps like Duolingo built their core loop around translation: see a Spanish sentence, fill in the missing word, match it to English. That loop is useful for memorizing individual words and basic grammar, but it rehearses the exact step that keeps readers translating. Every exercise confirms that the way to prove you understood a Spanish sentence is to produce or recognize its English counterpart. Nothing in that loop asks a learner to sit with meaning in Spanish and stay there.
This is why a long streak on a gap-fill app and an inability to read an actual paragraph of Spanish without translating it line by line can coexist in the same learner. The exercises were never testing direct reading. They were testing translation speed, which is a different skill with a different ceiling.
Retraining the habit
Four changes make the difference, and none of them require giving up the words you don't know yet.
Read below your frustration point
If you're stopping every third word, you're decoding, not reading, and decoding is where translation habits form. Dropping to a level where most of the page is already familiar is what lets the brain process Spanish as Spanish instead of as a code to crack.
Look up words without leaving the sentence
Switching to a separate dictionary app, typing the word, reading the entry, then finding your place again breaks whatever direct understanding had already built up. Checking a word's meaning in place keeps the sentence intact in your head instead of resetting it.
Reread before moving on
Translating usually happens on the first pass through a sentence, while you're still figuring out what it says. A second pass, once you already know the meaning, is where the brain gets a chance to process the Spanish directly instead of reconstructing it from an English version built a moment earlier.
Give a stop a few seconds, not thirty
The point of looking up a word is to keep moving, not to confirm every guess before continuing. A long pause to work out a sentence grammatically is useful occasionally, but as a habit it trains the same slow, puzzle-solving mode that translating does. A quick look, then straight back into the sentence, keeps the reading itself as the main event.
What this looks like in practice
In HabloHack, stories are graded to CEFR levels from A1 to C2, so reading happens where most of the page is already familiar instead of fighting unknown words on every line, which is what Nation's 98% threshold is actually asking for. Tapping a word shows its meaning, pronunciation, and an example sentence without leaving the story, so checking an unknown word costs a few seconds instead of a trip to another app and a lost place in the paragraph. Words looked up or saved go into a personal list alongside official and community lists, and flashcard review runs on FSRS, a spaced repetition system that calculates when each word is actually due instead of repeating everything on the same fixed schedule.
None of this is a shortcut around learning vocabulary. It's a way of practicing reading that matches how reading actually works, in Spanish or any other language: material you mostly already understand, a quick way to fill the gaps, and enough repetition at the right interval that the gaps close. That's the setup behind HabloHack, for readers who are done translating and ready to just read.