Listen to this clip in its entirety:
Now listen to the original sentence again:
Assuming you fall within a reasonable part of the bell curve of humans, you heard the woman abruptly sing “sometimes behave so strangely” in the middle of the phrase. Listen to it again…and again…you cannot escape the singsong interpretation of those words now.
Diana Deutsch of the University of California, San Diego first documented this phenomenon — the spontaneous translation by our brains of a spoken phrase to song when isolated and repeatedly looped — in a 1995 research study. Dr. Deutsch played a looped version of the phrase for eleven choir members and then asked them to reproduce what they heard. Here are their recollections:
Playing the sentence just once for another set of singers generated these reproductions:
This curiosity is not purely academic, as modern composers have been exploiting the effect for decades in their music. Steve Reich is probably the most notable example with pieces like “Different Trains” that derive much of their melodic content from the natural rhythms and pitch variations in speech recordings: