Your brain is too efficient – and that's the problem
Babies can tell apart every sound in every language on earth. At 12 months it's over: the brain cleans up and keeps only what's useful in the mother tongue. Japanese speakers lose R/L. German speakers learning Spanish lose the difference between a tapped R and a rolled RR. Not a defect – pure optimisation.
The mean part: what you don't hear, you can't say. You produce "pero" thinking it's "perro". Nobody corrects you because context usually makes it clear.
Minimal pairs examples: two words, one sound apart
A minimal pair is two words that differ in just one sound:
- pero (but) vs. perro (dog)
- caro (expensive) vs. carro (car)
- vino (wine) vs. fino (fine)
Two words, a tiny sound difference – only visible if your ear actually hears it.
Ear training: how to train your ears in 10 minutes a day
- Listen phase: One of the two words plays at random. You guess. At first you'll be 50 % right – pure chance.
- Feedback: Find out immediately if you were right. The brain only learns with instant feedback.
- Repetition: 5–10 minutes daily for 2–3 weeks. Suddenly it "clicks".
Reducing your foreign accent: which pairs first?
Don't drill what your mother tongue already covers. Drill what you never heard. For English speakers learning Spanish, in order:
- Tap R vs. rolled RR (highest priority – "pero" vs. "perro")
- Spanish D between vowels (sounds like soft "th": "nada" → "natha")
- The five clean Spanish vowels (very different from English's 12+)
Common questions
"Isn't more speaking enough?" No. Speaking without ear training fossilises your mistakes. You produce them wrong day after day – and your brain stores that as "right".
"When do I start speaking?" After 50–100 hours of focused ear training. By then you have the phonetic foundation. Before that: shadow, repeat, no free talking.