WHY RIYAAZ? - Page 8

Originally posted by: qwertyasdfgh

Sure but one thing is sure Riyaaz and creativity are unrelated and riyaaz doesnt give a jump-start for creativity either... Music is math...and tonal instinct is all thats necessary..

Riyaz has only bearing on attaining precision on tonal and microtonal level... there is no creativity.. attaining proficiency at that level doesnt necessarily get you to the door way of creativity...

very well put! bit to the "right" of the position i had in my long post, but very well articulated.

now, let's get to some of the issues:

1. interviews. these make for good light-reading for the most part and do not really offer path-breaking prescriptions one can really take to the bank. for every sonu nigam that made it big, there are a thousand others who started out on the same tracks and failed. and it is is not because they are necessarily stupid or unskilled or non-riyaazed. they might even be better, but they dont hold the "lottery-ticket". if you ask them after their failures, they'll tell u that none of the riyaaz etc paid off.

the conclusions derived from such interviews do not generally hold cross-sectionally for other people. sorry, am just pointing this out because we had pointed out elsewhere how someone's thinking in some areas were irrelevant to how they would think in a correlated area. pls realize that in that case we were talking about the same person, here we are exterpolating the unique, often lucky-break, experiences of one person onto others!

2. surtaal, the point you are making with the father-child failure stories is critically flawed. consider this: there are thousands of folks whose fathers are not singers, assume 5 of those kids make it. now assume 5 out of 20 star-kids makes it. given this, which of the two has higher odds of making it? get my drift? dont look at the absolute numbers, look at the probabilities. you are already looking at a very small pool of people when you are looking at stars, to then find even a few of their kids making it is actually pretty good.

did u understand the posts about the "accidents", "lottery tickets", dim-wit winners who make it 10 years in a row? some of my arguments are based around those, so it might help to refute them.

otherwise, i thought the post was a restatement of your earlier opinions, and do not address any of the elements in the intervening discussion.

bottom-line, nothing new. even the "watering down" of the riyaaz in later years is getting to the position i have had. not patronizing, but how i see it. it certainly does not seem that for someone accomplished to continue with a 16 hour schedule is worth it any more.

but what we do have now is an ideal guru, versatile and accomplished in different matters to be able to appropriately guide their shishyas. forget the 21st century. did we have enough of them around even at the start of the 20th century? to suggest a few names does not an education system make. and pls note, i am not questioning issues of benovolence or selflessness on the part of the gurus, i am just questioning versatility in different subject areas.

sorry that's how i saw it.

Originally posted by: chatbuster

yaar, another thing, in any large sample size, we can be sure to find "accidents". consider the following example- someone can be totally clueless about stocks, having just a 50% probabibility of winning or losing money any year. yet, he/ she has a 1 divided by 2 to the power of 10 probability of making money 10 years in a row. in a population of just 10,000 such dim-wits, we'd then expect to find 10 people like him who make money every year. so is he really smart? no! he is just one of the 10 lottery-ticket holders we'd expect to have in a sample of 10,000.

ditto perhaps for the riyaaz folks. a thousand people start off on the riyaaz tracks every year. a few make it. an accident waiting to happen? lucky lottery winners?

Now the above rationale is too simplistic and the logic cannot be extended to our discussion. I feel you are saying that those with Riyaaz are lucky lottery winners. Would it not apply to those that do not perform Riyaaz??? In such a case, every successful or an unsuccessful person has been waiting for an accident to happen.

In quantitative terms, I can draw a linear graph with 2 variables holding all other things constant.

Success = C1*x + C2*y + C3
x = Measure Riyaaz, y = Time
C1, C2 are constants (combining legacy, destiny, luck, talent etc..)
C3 = Creativity
If C3 were large enough x and y become irrelevant.

Or to go down the road you suggest . . .

Statistically speaking there would be outliers in any normal sample size. If we were to consider those that fall in the bell curve, and borrow from your analogy of a lottery and apply to the participants in reality shows like SRGMP or II. The contestants go through the elimination process where we would not get to see any sons or daughters of established musicians/ performers. So your legacy argument does not hold water or let us say it is a level playing field.

So we are now left with those that either have natural talent or have acquired the skills through training. You have been witness to their performances and have gotten to judge them. So the sample size is reducing or let us say we are eliminating good performers from the better ones. It is an iterative process sort of like mathematical analysis.

There are those that have not had the rigourouse training like Abhijit Sawant (II' 05), Vinit Singh (runnersup SRGMP)., no legacy whatsoever . . . we have seen their parents on interviews. Since you had brought up the probability theory that those with legacy have better chances of being successful. But they do not go through any lottery. You cannot call them standing in a queue for an accident to happen, they may or maynot be successful. But they get to take a shot without standing in line. All those that participated and were, say in the final 10 or 20 are capable of going to the next level if they now use their creativity to build on top, since they now have visibility in some way close to what the celebtrity kids have. We are now trying level the playing field further. You cannot classify these as dim wit winners because on a level playing field (all things considered equal) everybody has an equal chance of winning, but it is not a game of chance alone.

If you take this small group and say they belong to a huge sample and is a small percentage, you cannot win a case against India's population. Also we cannot extrapolate this to the original population. However, we can establish a positive one-way correlation between Riyaaz & Creativity. (meaning Riyaaz inspires Creativity positively, not negatively) not the other way around. Just as all the sportsmen have coaches, the music aspirants should have Gurus to guide and coach them

Our inadequate training methods and other considerations prevent us from winning Gold Medals in Olymics. That should be a different thread.

i think the problem here is one of "moving target". i seem to be running up against different definitions of riyaaz, of success, of creativity, depending on the person on the other side. what specifically is your working definition of riyaaz? is it punjini's or the other "evolved" one as a catch-all for experimentation, learning and drill? also, notice your previous post- you were talking about riyaaz versus creativity, now you are talking about riyaaz versus success. also, there seems to be imprecision with regard to success, though not with your post. from the example of various personalities quoted, i think the reference is to "enormous success" but i cant be sure. finally, learning and experimentation. more on this below.

rather than work with all the imprecision, let me just state my position. first, my position is that music education/ learning is positively correlated with creativity. learning for me is life-long. it goes far beyond drill-riyaaz, although in the initial stages of learning, some drill-type riyaaz is necessary to acquire basic proficiency. second, excessive drill-type riyaaz is bad if you want creativity. third, some level of proficiency is needed for success in any field. beyond that, we need experimentation to find our own "voice", our own soul.

now let's get back to the father-son legacy issues you bring up. well, that's a perfect case to illustrate my "odds" theory. 5 out of 20 star-sons make it, 5 out of 1000 non star-sons make it. who has the higher odds of success? obviously, the star-son. and i am holding everything else equal, the talent, the creativity etc. Actually, it's a very easy point to make and am surprised we still dont get it.

now, what of the sgmp example you brought up. well, we dont have any star-son data to prove or disprove the existence of the star-son factor in sgmp, do we? a statistician wld try and work with the data that does exist. and we do have that data in the real life beyond sgmp. the "odds" example above is based on that. of course, the 5/20 and the 5/1000 numbers are made up, but i firmly believe that that point would hold up if we crunched out the numbers.

as for your linear regression equation, firstly that's the wrong relationship to be looking at. in my mind, the relationship is non-linear, not linear. at best, we can expect to have some positive relationship at low levels of riyaaz. after a point, i wld expect the relationship to turn negative for reasons of muscle-depletion, burn-out, time taken away from more useful experimentation.

the choice of the dependent variable is further very important in any case. if we use "enormous success" as the criteria, then the regression formulation would still "prove" my point. btw, i wld not have used the regression formulation on intellectuall grounds because i do find structural problems with it, but am doing so only because you asked for it. well, why wld it prove my point? simply because "enormous success" is rare. we shld then find a significant t-statistic for the father-son relnship, very little for riyaaz. why very little for riyaaz? simply because "enormous success" is so rare that even when we had riyaaz, we wld find the vast majority of cases to show a 0 success factor. note that the appropriate value to use is t-stat, not the coefficient when proving/ disproving relationships. in any case, i wld also expect to find a large unexplained r-square, showing the existence of factors we can lump under "luck".

sorry for the long one but i hope you got the point...

Very interesting !!
For the data to yield meaningful results in regression we need to have a normal distribution. The moment you talk about Star-sons and military-style riyaaz, we start looking at outliers.

As far as regression is concernde choice of variables is indeed very important (maybe most important once we clean out the data).

It appears to be a volley game between us now. However, I am happy that we are in agreement on atleast some of the aspects

Originally posted by: chatbuster

rather than work with all the imprecision, let me just state my position. first, my position is that music education/ learning is positively correlated with creativity. learning for me is life-long. it goes far beyond drill-riyaaz, although in the initial stages of learning, some drill-type riyaaz is necessary to acquire basic proficiency. second, excessive drill-type riyaaz is bad if you want creativity. third, some level of proficiency is needed for success in any field. beyond that, we need experimentation to find our own "voice", our own soul.

Now statistically we can, based on how we manipulate data, prove and disprove the Star-son/ daughter (to be politically correct) sucees to be more or less than Non-Star son/ daughter success.

Let us not quantify success yet.

Originally posted by: chatbuster

now let's get back to the father-son legacy issues you bring up. well, that's a perfect case to illustrate my "odds" theory. 5 out of 20 star-sons make it, 5 out of 1000 non star-sons make it. who has the higher odds of success? obviously, the star-son. and i am holding everything else equal, the talent, the creativity etc. Actually, it's a very easy point to make and am surprised we still dont get it.

No surprise, we seem to holding ground and working with different datasets and assumptions.

Originally posted by: chatbuster

now, what of the sgmp example you brought up. well, we dont have any star-son data to prove or disprove the existence of the star-son factor in sgmp, do we? a statistician wld try and work with the data that does exist. and we do have that data in the real life beyond sgmp. the "odds" example above is based on that. of course, the 5/20 and the 5/1000 numbers are made up, but i firmly believe that that point would hold up if we crunched out the numbers.

If we use 5/20 and 5/1000 Non-Star sons/daughters look very bad.
But it is unfortunately for your data is not 5. Almost all the entrants into art and music are guys with no-celeb legacy. but again, your data set is yours.

Originally posted by: chatbuster

as for your linear regression equation, firstly that's the wrong relationship to be looking at. in my mind, the relationship is non-linear, not linear. at best, we can expect to have some positive relationship at low levels of riyaaz. after a point, i wld expect the relationship to turn negative for reasons of muscle-depletion, burn-out, time taken away from more useful experimentation.

This is unreasonable Riyaaz. Nobody does that. Just for fun why dont you take some Hindustani/ Carnatic music classes. A lot of myth will clear out for you.

Originally posted by: chatbuster

the choice of the dependent variable is further very important in any case. I agree if we use "enormous success" as the criteria would not like to quantify success yet, then the regression formulation would NOT still "prove" my point. btw, i wld not have used the regression formulation on intellectuall grounds because i do find structural problems with it, but am doing so only because you asked for it. well, why wld it prove my point? simply because "enormous success" is rare. we shld then find a significant t-statistic for the father-son relnship (more than 95% of the entrants are not celebs, which proves my point of Legacy-independence and it meets your t-stat), very little for riyaaz. why very little for riyaaz? simply because "enormous success" is so rare that even when we had riyaaz, we wld find the vast majority of cases to show a 0 success factor. note that the appropriate value to use is t-stat, not the coefficient when proving/ disproving relationships. in any case, i wld also expect to find a large unexplained r-square (sure !!! and we would not know how each one is/ is not correlated), showing the existence of factors we can lump under "luck".

sorry for the long one but i hope you got the point... (back at you, buddy ;)😉

am very saddened to learn of sweetdisha's passing away. knowing how much she loved classical music and her admiration for AR Rehman, as a mark of respect and affection for her, i would like to end the debate from my side for now... For me, it was an intellectual excercize, for others i guess it meant a whole lot more...

Even I am unable to think straight after hearing about the passing away of a member who was making posts until yesterday! Will read all the arguments when I am in the right frame of mind. Though I feel that whatever needs to be said on the subject of riyaaz has been said.

I had seen her post on Hindustani Raagas just yesterday.
It is very unfortunate that she passed on!

I have not had the opportunity to exchange thoughts with her. May her soul rest in peace.

Originally posted by: Surtaal

Very interesting !!
For the data to yield meaningful results in regression we need to have a normal distribution. The moment you talk about Star-sons and military-style riyaaz, we start looking at outliers.

As far as regression is concernde choice of variables is indeed very important (maybe most important once we clean out the data).

It appears to be a volley game between us now. However, I am happy that we are in agreement on atleast some of the aspects

Now statistically we can, based on how we manipulate data, prove and disprove the Star-son/ daughter (to be politically correct) sucees to be more or less than Non-Star son/ daughter success.

Let us not quantify success yet.

No surprise, we seem to holding ground and working with different datasets and assumptions.

If we use 5/20 and 5/1000 Non-Star sons/daughters look very bad.
But it is unfortunately for your data is not 5. Almost all the entrants into art and music are guys with no-celeb legacy. but again, your data set is yours.

This is unreasonable Riyaaz. Nobody does that. Just for fun why dont you take some Hindustani/ Carnatic music classes. A lot of myth will clear out for you.

forgive me, cld not resist! yaar, you are talking about sgmp entrants? come on. are those the examples of enormous success you cld find? there are real-life examples, from KK's son to Rafi's son to Mukesh' son, to Zakir Hussain. How many "great successes" have we had? 5, 10 or 50 at most, depending on how loosely one defines enormous success? and we've had a few of their sons at least make it fairly big, even if not all the way. that's a pretty good ratio. dont even need precise data to see the point. and how many others without star fathers have tried and failed? thousands. maybe millions.

and yes, your equation does reflect your belief that more riyaaz the better. so why get queasy now with excessive riyaaz? and from the way folks have talked about hard riyaaz, it wld seem that they are talking about "a whole lot of riyaaz". in any case, if you are now watering that down, then fine. we all do some amt of minimalistic riyaaz in one form or the other in our fields.

and again u are being imprecise when u talk about normal distribution. the errors have to be IID normally distributed. and that is why i pointed out the flaw with the linear regression when u brought it up in the first place. cant win either way. u bring it up, i point the flaw in it, but continue with your example, then u suggest we shldn't be using it. take your pick.

on other aspects, there is at best a weak-to-medium relationship between education and creativity (i might be able to dig up a link to that research, but u can find it on ur own with google). with riyaaz, we are not even talking about education. we are talking abt drill. i know we now want to run away from multiplication tables and distance ourselves from that analogy, but it is a good one to use to illustrate how absurd it wld be to correlate drill/ tables with creativity. certainly u wld not find the level of creativity that we seem to have been attributing to riyaaz.

cld we pick this up after a few days... given what happened? i really dont want to be indulging in this today.

Nice post punji ... I just wish you could say all this two the EMET participants.