so strangely experiments

[William Hartmann <hartmann@xxxxxxxxxx>]

Dear List,

Here are a few informal experimental results on "... sometimes behave so
strangely."

1. There is a tendency for repetition to bring out the melody hiding in 
any spoken phrase. It is possible to enhance the effect of an emergent 
melody by speaking tunefully, without actually singing. It is also 
possible to prevent any melody from emerging by including a glide on F0 
in each syllable. The glide does not need to be exaggerated, but 
exaggeration ensures that no melody emerges -- not even a fragment.

2. Speech tones that happen to hit close to the notes of a scale seem to 
lead to a more salient emergent melody.

3. It's a common experience that once one has heard the emergent melody, 
one cannot listen to the spoken phrase without hearing that melody. This 
reminded me of ear worm melodies that go around and around in one's head 
and won't go away.  I've found that the only effective way to remove an 
ear worm is to replace it by a different ear worm.

And so it seemed possible that an emergent melody might be eliminated by
replacing it by a different emergent melody.  To test this idea, I used 
the same words with the same rhythm but changed the direction of the F0 
contour for the two syllables in "sometimes" and in "strangely." The new 
melody was caused to emerge by repetition. However, the new melody did 
not remove the old melody. Each spoken phrase continued to cause its own 
melody to emerge.

Best wishes to all for the holidays,

Bill Hartmann