Marc Quinn's Preserved Garden.

While I probably agree with the findings that they summarize in this story, it seems to me that it's not the best test. For example, the music that was deemed "angry" went into the lower musical registers. So that may just mimic an angry person's voice, and we get our cue from that. The "peaceful" music was slow. I think the study needs to be widened. What about slow music that's very dissonant? Probably not peaceful. What about angry music that's high in pitch? Maybe not angry anymore. I think they're making some assumptions that aren't necessarily completely true, and leaving out some of the possibilities that cause people to interpret music emotionally. Rhythm, pitch, instrument timbre, arrangement, orchestration. Music has so many variables, and this seems to be far from being able to make such definitive statements as it does.