Music is surprisingly mysterious, for something so ubiquitous. For example, it’s not really clear why we generally associate major keys with happy moods and minor keys with more somber feelings. Also, we choose our scales somewhat arbitrarily out of a range of possibilities. Within a single octave, humans can discern about 240 different musical tones, but the ways we divide this complex tonal landscape are fairly uniform across not only western music but at least some other musical traditions, despite the multitude of other options.
A couple of papers from the lab of Dale Purves at Duke suggest that the answers to both questions are linked to the properties of human speech.
A paper in the Journal of the Acoustical Society of America reports on research comparing the tonal qualities of excited and subdued speech and finds that the former contains more major intervals and the latter more minor intervals, which suggests a source for our identification of the emotional qualities of music in major and minor keys. Another paper in PLoS One shows that the musical intervals making up the most widely used scales are those that most closely resemble the harmonic structure of vowel sounds appearing in human speech.
These close links between the tonal qualities of music and speech suggest that one reason music is such a powerful influence on humans is that it uses whatever mental machinery evolved to pay attention to the utterances of other humans (or as the Purves lab web page puts it, “These findings are consistent with the idea that humans have a bias for conspecific vocalizations.”).
You can read an article from Science Daily about this work. The two papers are:
Major and Minor Music Compared to Excited and Subdued Speech, by D.L. Bowling, K. Gill, J.D. Choi, J. Prinz, D. Purves. Journal of the Acoustical Society of America, December 2009.
A Biological Rationale for Musical Scales, by K. Gill and D. Purves.
PLoS One, 4(12): e8144. doi:10.1371/journal.pone.0008144, published December 3, 2009.