When we listen to improvisational jazz, or solo classical violinists, the way they phrase and inflect melodies feels vocal, like they're talking to us. When I was figuring out how to perform solo, I wanted to move back and forth between bass riffs, melody, and harmony, so I often used sounds instead of - or alongside - the words of a song. I found that if I sang a line using the consonants, vowels, shadings, and inflection we recognize as human language sounds, people responded as if I were talking to them.