The research behind learning a language by reading
LingoBlend is built on a handful of well-documented findings from language acquisition and cognitive psychology. Here is what the research actually says, and where each idea shows up in the app.
The diglot weave: reading in two languages at once
In 1968, the linguist Robbins Burling described a teaching technique he called the diglot weave: take a story in a language the reader knows, swap a small share of its words for their foreign-language equivalents, then raise that share as the reader grows comfortable. Each new word is understood from the sentence around it, so the reader picks it up without stopping to translate.
Smart Blend is an automated, modern version of that idea. Paste any article, book chapter, or web page, move the slider anywhere from 10 to 80 percent, and the app blends that portion of the text into the language you are learning. Tap a blended word to see its translation, its grammar context — tense, conjugation, and base form — and save it to your dictionary in one motion.
Comprehensible input you can dial in
Stephen Krashen's input hypothesis holds that we acquire a language when we understand messages pitched just beyond our current level, what he labeled i+1. The practical obstacle has always been supply: authentic text is rarely written at exactly the level you need today.
A blend percentage solves that. Start low when a language is new and raise it as you gain ground, so the challenge tracks your progress instead of a fixed textbook grade. Because the surrounding words stay in a language you already read fluently, each sentence remains understandable even when a target word is unfamiliar, and its meaning arrives from context the way it does in real immersion.
Getting words to stay
Reading gets vocabulary in; a few well-studied mechanisms keep it there. LingoBlend schedules reviews with an SM-2 algorithm, the same spaced-repetition family behind Anki: a new word returns after 10 minutes, an hour, and eight hours, then graduates to intervals of 1, 3, 7, 14, 30, 60, and 120 days as you keep answering it correctly. Each review lands close to the moment you would otherwise forget, which is far more efficient than massed cramming.
Two encoding techniques make each word easier to retrieve. Dual coding (Paivio, 1971) says a word stored alongside a picture leaves two memory traces, verbal and visual, that reinforce each other, so you can attach an image and a personal memory trick to any saved word. The keyword method (Atkinson & Raugh, 1975) links a new word to a familiar sound or a vivid mental image for deeper encoding. The five practice games, plus the Pro Word Story that spins a short narrated tale out of your own saved words, keep returning those words to you in fresh contexts.
Why blending works: the science
Text blending isn't just clever — it's backed by decades of language acquisition research.
Comprehensible Input (i+1)
Language is acquired when learners receive input slightly above their current level. The blend slider gives you adjustable i+1 in real time — always the right amount of challenge.
Code-Switching & Bilingual Processing
Research shows that mixed-language text captures greater attention and enhances vocabulary acquisition. Blending mirrors how bilinguals naturally process language.
Incidental Vocabulary Acquisition
Learners acquire substantial vocabulary naturally through exposure to words in informative contexts — not through isolated memorization. Blended text makes this happen effortlessly.
And every other feature is backed by science too
Spaced Repetition
SM-2 algorithm in all 6 games schedules reviews at optimal intervals — up to 25% higher retention than massed study.
Active Recall
Games force retrieval from memory, producing 51% better retention compared to passive review.
Contextual Learning
Words encountered in rich, varied contexts through stories and sentences are retained significantly better.
Dual Coding Theory
Words paired with images create two memory traces instead of one, making recall 2-3× stronger and more durable.
Keyword Method
Linking new words to familiar sounds or personal stories creates deep encoding — 1.5× better recall than rote memorization.
Narrative Context
Words encountered in stories are retained significantly better than isolated lists — narrative provides meaning, context, and emotional anchoring.
Start learning a new language today
Join LingoBlend and turn any text into a personalized language lesson. Free to start, no credit card required.