Some of you will remember a few weeks back I said I was planning to do research regarding popular song keys and the color of sound so to speak. Well in my research I discovered a company out of Barcelona which has done far more than I could attempt to achieve. They have studied 3.5 MILLION popular songs from the 1950s forward.

Each one of the songs appeared as a little white dot and these dots seem to form clusters, or rather hit song constellations. There were 7 when I read the articles, dated a few years ago. How this affects you as a songwriter or publisher is this, they will pop your song or songs into their computer program (for a fee of course) and if your song appears (as a red dot) within one of those "hit song" clusters then your song has similar components to previous hit songs and therefore a good rating for success. If your song appears in a far off galaxy they suggest you will have a harder climb to "stardom", pun intended.

They initially scanned Nora Jones, prior to her great success and found 8 songs from her CD landed within those clusters. I was rather surprised that this technology emerged 3 years ago and was on a few news stations in the States, but the company's name has never surfaced on any music related website I frequent. Nor when I asked my associates if anyone had done this sort of research, when initially undertaking my own, no one made mention of this service. So when I found this site three weeks ago I was rather excited by its concept and wanted to say to everyone "check THIS out"…but I refrained and thought I would wait until now.

They call themselves "Hit Song Science" or "Polyphonic Human Media Interface". Apparently humans do prefer particular patterns of music and melody… perhaps the number 7 relates to 7 notes? I had questions! I wanted to know tempos, keys, time signatures etc. but when I contacted Roger Cortal Commercial Department Director - Spain he could only offer "We could say that each cluster represents a musical - mathematical formula, being a specific combination of rhythms, structures, harmonies, beats, tempo, etc....". I wanted to know if these 7 clusters were language specific or do they cross all cultures as I knew they input songs from Billboard charts. "We analyze the song for a given market. So it could be that a song sang in Japanese becomes a hit in the German market, for example". Which lead to the question of whether or not the music of various cultures landed in certain clusters or do they cross into all clusters. For instance if we numbered the clusters from 1 to 7 do Spanish songs primarily land in 3 and 5 for example, and English songs in 1 and 3 and 6 or does it matter? "Each market has different specific numbers of clusters. For the moment we have discovered more than 40 clusters, within all the markets" explains Roger, "and let's say that the Spanish market has 10 hit clusters. Not all 10 of them are activated every year. We can see them when they are activated". Not the clearest of answers but it made me think if they are not all activated each year, then are there 'seasons' or 'rotations' of musical tastes to which he replied "Yes, it shows a trend in musical tastes". Ahhh seasons in the song!

They won't share with us, exactly what features make a great song, only that their system can tell you if your songs contain those features. Some of you may think it is a load of hooey and some may think it makes sense. I'm one of those. If you ask any publisher or A&R person, they will tell you, when they listen to a song they ask themselves "does it sound like a hit". This technology in my eyes only confirms to me, that hits became hits because on some subtle levels a large percentage of our populace responded to them . The term popular music is just that: popular, because the majority of the population enjoys the song. At this point it is still a guessing game as to the exact components of songs that humans respond to more favorably and I know I can't survey 3.5 million tunes.

You can visit them at www.hitsongscience.com. They will have a new website up and running in two weeks.

Earlier this year there was a news story I read on Reuters where 14,000 students were asked to listen to and rate music from a number of unknown artists. There were subgroups, some, separate from their peers, and the others who could view which songs others were downloading. In the blind group, uninfluenced by peer groups, the results were that the worst songs were given an "OK" and the best songs were given an "OK". Researchers drew one conclusion from this and that was that teenagers have no confidence in their ability to judge music and rely heavily on marketing and peer pressure. This was proven by the increase in particular downloads in the other group because simply, some kids would download songs because others already had.

Study co-author Matthew Salganik a sociologist at Columbia University suggests we are faced with an overload of songs, and a "desire for compatibility with others could drive the (music) choice since much of the pleasure from listening to music and reading books stems from discussing them with friends"*.

So this study suggests the reason songs are popular is not because we respond to them on a deep and spiritual resonance but simply because we prefer to like what our friends like. Hmmm. I don't agree it's that simple, maybe age is a factor there. Did they study people over 25 or 30 or 40? Personally, I love the way songs can make me feel. I can sit and listen at a songwriters circle and be moved to tears by a song I have never heard before. No peer pressure there!
Presently there is a study being conducted by Columbia Music Lab if you would like to participate please do so at http://musiclab.columbia.edu/general.html and let them see that some of us, over the age of 18, can actually like or dislike a song based on the song's merit, its musical structure and our response to it! Hopefully we will gain more insight into why and how we respond to music.

My quest continues.

~ Jana Reid