Preliminary hypotheses on music...

UPDATE:
Here is an amazingly insightful article on music theory, one of the best ever i have come across online. Something that every carnatic musician should know, which i doubt they do.
Hats off Srikanth.

(Sincere thanks to "N.Vijayanand")
The first part of this post is a verbatim reproduction of Vijayanand's thoughts on his blog, reproduced here for the lazy reader.
==========================
I always had this view that composing songs that were bad was a really hard task. Because of the simple fact that we "get used" to the song and start liking it after repeated listenings.

So what exactly do we mean when we say a partticular song/music is good.
It can only be governed by the following function H(t) : The 'happiness' (due to lack of a better word) we
derive from the song/music as a function of time. Four parameters for this function govern the quality of the song. The rate at which the function reaches its maximum value, the maximum value, the rate at which the function decays from the maximum value and finally "hysterisis",the rate of slow trace back(increase of happiness function) of the curve due to long periods of not listening to the song.

Let us look at the parameters individually:

1) Attack : (rate of increase): This , I guess, is dependant on the level of musical expertise of the person.
For example consider the song "Azhagana Ratchasiye.." -- If this song is heard by a person who is familiar with the raaga reetigowla, then the the attack will be higher for that person. The next obvious qustion is why? I think the answer has to deal with psychological reasons which I would reserve for a future post.

2) Decay (rate of decrease): After listening to a song repeatedly, we tend to lose interest in the song. Again I think the reasons for this are psychological. I think this phenomenon may be independent of the musical expertise of the person.

3) Maximum value: How good a song is should depend on the maximum value. But based on my personal experience it is very hard to distingush the maximum values of two good songs/music. (I get the same "kick" from all songs when I am enjoying it the greatest!)

4) Hysterisis: Suppose we hear a song continually for a week. By this time, the H value would have increased to its maximum value and would have started falling say to value 'l' (say) after a week. Then
if we we wait for a year and again hear the song. Now the H value will not be 'l' , but a much higher value beacuse of the trace back in the periods of non-listening.

Now let us explain the reasons for the following using our theory:

i) Common men enjoy film music more than classical music.
Based on two reasons:
a) Musical expertise of common man is low. So often the attack (rate at which H(t) increases) can be low.
b) Common man listens to (is forced to listen) to film songs in the initial stages, (even when the H for that songs has not risen).

ii) The old melodies dont die, the newer songs die a quick death:
This is a fallacy. Again explained using the frequencies of listening and hysterisis. (The old songs are heard at a mush lesser frequency compared to new songs).

iii) Carnatic music is of inherently higher quality than film music
I would have say no and yes to this.
I would say inherently there is no difference in the quality of music. (All the four parameters are probably not much better than film music)
But the reason for carnatic music's speciality is "manodharma" (the creative part, or the improvising part). By having this we have ensured we are not hearing the same (ditto) song/music again and again.
Each time we are hearing an improvisation. That is the reason ragas and krithis dont decay at all.
=====================
On top of that analysis, let me also say that there are 2 factors which influence the "happiness quotient" of a particular song.

a. Singability (simplicity to the ear)
b. Singalongability

Singability - the non-musician's perspective:
is a measure of how close to the original can a person reproduce the song or the effect it has on a listener. am talkin about non-musicians here, who value the singing component of a song generally much higher than the orchestration (save for a few explicit pieces of backing).

Say, for example, a song like Sa re ga me (of BOYS) fame, though an amazing song, has a low value for singability, simply because the orchestration is such an integral part of the song. Sans the orchestration, the song cannot be sung alone.

The "fallacy" that Vijayanand talks about old/new songs is accounted for here, again for the simple reason that a greater emphasis on orchestration and technical finesse in newer songs (from the point of view of "sounding" good - with nothing mentioned about the musicality here) sort of distinguishes old songs(i am not talking about the intensely classical "oru naall podhuma" or the "paattum naane" types) as some that you would easily relate to (being a non-musician).
These are the songs that have a large hummability value. At any random public function, or at a gathering, these are the songs, the singable ones, that one chooses to sing, because most of the essential flavor is retained.

Lets look at a few examples of GOOD songs with high and low singability:

Nilave ennidam nerungaadhe : Old with high singability
Konjum nilavu : New with low singability
Machaan peru madhura: New with High singability
Oru naall podhuma: Old number with low singability
Carnatic music: in general, low singability.

2. Singalongability (the musician's perspective):
Arbitly named, this is a measure of the same from a musician's perspective. When performing a song, the technical- and technological- factors in the song that render it easy or difficult to perform.
Say, a song like konjum nilavu (shown above to have low singability) has a high singalongability, and thus, a musician, with the 'band', would love to perform this song, and the viewers (non-musicians) would go crazy "singing along".

These songs would be sung infrequently on occasions, and this is no way erodes their "lifespan".
All singable songs i guess are also highly singalongable.
The converse is not always true.

GOOD Songs with low values of both might be best admired with the film... or heard over and over, but never "sung".
eg: Enakkoru Girlfriend venumadaa. Too much of technology in the song. Too many layers. Lots of vocal harmony. Even ARR can't perform it live on stage without resorting to a CD playing a (-1) track.

I know i should put some more thought into this and streamlined them, to actually speak some proper sense, but i never seem to have the patience these days..
Improvements, corrollaries, flaws... please do voice it out.
This could be interesting.
Just my 2 paise.

Written & Directed by,
Girish.