Hearing Sounds That Aren't There
In 1970, psychologist Richard Warren made a remarkable discovery: when a sound in a word is completely replaced by noise (like a cough), listeners still hear the missing sound.
Not "figure out" what it should be. Not "guess." They genuinely perceive the sound as being there.
In Warren's original experiment, 19 out of 20 subjects couldn't even identify which sound had been replaced!
The 's' in "legislatures" was completely replaced by a cough sound.
Result: Listeners reported hearing the complete word "legislatures" with a cough somewhere nearby—but not replacing any sound!
While we can't perfectly replicate audio online, we can demonstrate the same principle with text. Your brain uses context to fill in missing information automatically.
What word appeared where the orange block is?
Watch the waveform. The orange section represents noise replacing a speech sound.
The phonemic restoration effect reveals that speech perception is not just bottom-up (ears → brain) but heavily top-down (brain → perception).
Your brain uses:
The result: you literally hear sounds that aren't in the audio signal!
When the missing sound is replaced by noise (cough, static, tone), restoration works powerfully.
The noise provides "cover" — your brain assumes the sound is there but masked.
When there's just silence, restoration fails. You notice the gap.
Silence signals "nothing was said" rather than "something was obscured."
This proves the brain isn't just "guessing" — it's actively constructing perception based on what should be there given the masking sound.
Direct recordings from human auditory cortex revealed something remarkable:
"The lateral superior temporal gyrus (STG) represents the missing sound that listeners perceive. Neural activity predicting what sound will be 'heard' appears up to 300 milliseconds BEFORE the sound is even presented."
Your brain is constantly predicting upcoming sounds and fills in the prediction when the actual sound is masked.