This is very cool. Listen to the first clip. Sounds like twitters and blips.
Then listen to the second clip.
Now go back and listen to the first. It becomes comprehensible once you know what you’re looking for.
This is called ‘sine-wave speech’. When a linguist records your speech using a spectrometer, there are dark patches of high intensity, called formants. Draw the formants using sine waves, and you get the twittery sound that resembles somewhat-but-not-quite speech.
I guess this is yet another example of how perception depends on the knowledge and expectations of the perceiver. Something to remember when I try and understand the voting habits of others.
17 November 2008 at 12:05 pm
Hmm I didn’t quite get it. I immediately recognised the former as clear speech with very minimal distortion. Perhaps if i didn’t know that I was meant to be looking for ‘something’.
17 November 2008 at 2:19 pm
I found the first one easily understandable, but the subsequent examples were harder. I would heard incomprehensible squeaks with a couple of words thrown in. Interesting.