Most popular language apps ask you to match cards in silence, tap answer options with a finger or fill in missing letters. Cognitive science has known for fifteen years that this way of taking information in is fundamentally weaker than saying the words aloud. The phenomenon has a name: the production effect.
The finding: Colin MacLeod's eight experiments
In 2010 a group of psychologists led by Professor Colin MacLeod (University of Waterloo, Canada) ran a series of eight detailed experiments to work out how the mechanical act of speaking affects long-term retention.
Participants were given lists of words. They read one part of the list silently, to themselves, and spoke the other part aloud in their own voice. Across every test of free recall and recognition, the words that had been spoken aloud were retained significantly better.
"The simple act of reading words aloud led to considerably better memory than reading them silently. Production is a simple but remarkably powerful way to improve human memory." MacLeod & Bodner, Current Directions in Psychological Science (2017)
Why it happens: three neurophysiological traces
When you read a foreign word silently, or tap a phone screen without speaking, the brain uses a single sensory channel — vision. It saves energy and processes the information shallowly.
Say the word aloud and a full sensorimotor loop switches on:
- Motor trace: the brain plans and sends commands to the muscles of the tongue, lips and vocal folds to articulate an unfamiliar sequence of sounds.
- Auditory feedback: your ears hear your own voice, laying down an acoustic imprint in the temporal cortex.
- Distinctiveness: a spoken word becomes a subjectively unique event in your own biography, standing out against the passive stream of thought.
What this changes about learning a language
For a second language (L2) the production effect matters enormously. In your native language the motor programme for every word has been polished to automaticity since childhood. In a new language those programmes simply do not exist yet.
If you merely "understand" a word with your eyes inside an app, you are not training the speech apparatus. Then a real situation arrives — a café, a hotel, or an airport — and you freeze: the brain recognises the word, but the lips cannot assemble it.
How the production effect is built into Lexy
Lexy's method is organised strictly around speech:
- Live voice dialogue without typing: during the core conversation stage with Lexy, you speak exclusively out loud — this fires up the brain's speech centers and breaks down the hesitation barrier.
- Immediate dialogue with the AI: Lexy speaks to you, you listen and reply on the spot in your own voice.
- Exercises and tests at a calm pace: after the intense voice dialogue, Stage 3 ("Exercises") lets you listen to phrases, type responses on the keyboard, and take comprehension quizzes. This is designed deliberately to develop spelling, grammar, and listening skills in a calm rhythm, protecting the learner from cognitive fatigue.
Test the speaking effect on yourself in lesson one
In Lexy the first two lessons of any of the 37 languages (such as Spanish for travel or English from scratch) are free, with no bank card. The first one opens with no sign-up — say your first phrases right now.