No Causal Effect of Music Practice on Ability (2014) [pdf]
gwern.net
gwern.net
The "10,000 hours to expert performance" notion originated in Ericsson 1993 (http://projects.ict.usc.edu/itw/gel/EricssonDeliberatePracti...) , which defines a very specific term, "deliberate practice" -- "In contrast to play, deliberate practice is a highly structured activity, the explicit goal of which is to improve performance. Specific tasks are invented to overcome weaknesses, and performance is carefully monitored to provide cues for ways to improve it further. We claim that deliberate practice requires effort and is not inherently enjoyable." Ericsson adds more detail in the full paper.
It's not just "are you playing piano for X hours per day".
The paper in this post mentions deliberate practice in its review of the literature, but then all of a sudden and without explanation reverts to the more generic term "practice" for all of its contributions, which appears to be defined as "how many hours did you play your instrument".
Note that "play" and "practice" are explicitly disambiguated in Ericsson.
I have a hard time believing that the authors are unaware of this distinction, especially considering that they heavily reference Ericsson.
Ericsson's original paper says: "In this article we propose a theoretical framework that explains expert performance in terms of acquired characteristics resulting from extended deliberate practice and that limits the role of innate (inherited) characteristics to general levels of activity and emotionality." They want to claim that the only innate talent that matters is that you are active and emotional, and not that you have any innate talent for music. This paper shows that the amount of time spent practicing appears to have no effect at all on some of these basic musical skills. It shows the presence of innate musical talent, something Ericsson wants to deny.
Some people really like to believe the 10,000 hour rule. I have no desire to talk someone out of believing it if they find it motivational. But as a musician, it's obvious to me that inborn talent is absolutely necessary for high-level performance. The idea that anybody can be Joshua Bell if they just put in the time is a wishful fantasy, not reality.
This kind of "blank slatism" is popular because it satisfies our desire for the world to be fair and just. We want to believe that we can create a perfectly fair playing field where success comes from effort alone. This is a noble idea, but the danger is when we discount evidence to the contrary -- evidence that sometimes people really are different, possibly having different talents or desires. Because then we look for who to blame when things don't end up the way we thought they should.
Practice matters. Especially at the upper level. There are hordes of talented "strong casuals" out there.
For becoming a chess GM you need 3 things: talent, grit(deliberate practice), starting early(10-12 is okay, 18 is too late these days).
One interesting aspect of starting early AND also persistent grit is being the youngest sibling in a chess family.
For example Hikaru had a master level older brother and of course there's Judith. The youngest siblings are ridiculously competitive against older ones.
No one with a musical background will be surprised that ear training (i.e the ability to hear musical details) is only distantly related to finger training, especially if the finger training happens on an instrument with discrete pitches, like a piano.
Ear training is a separate process, and has to be practiced separately. Playing will improve it a little, but won't fully develop it.
In fact amateur keyboard players are notorious for poor pitch discrimination. Keyboards are just switches. You don't have to find the pitch while listening for intonation, as you do on instruments that don't have frets and require very precise finger or mouth control, as well as the ability to listen to the rest of an ensemble.
http://www.vulture.com/2018/02/quincy-jones-in-conversation....
David Marchese: You’re talking about business not music, but, and I mean this respectfully, don’t some of your thoughts about music fall under the category of “back in my day”?
Quincy Jones: Musical principles exist, man. Musicians today can’t go all the way with the music because they haven’t done their homework with the left brain. Music is emotion and science. You don’t have to practice emotion because that comes naturally. Technique is different. If you can’t get your finger between three and four and seven and eight on a piano, you can’t play. You can only get so far without technique. People limit themselves musically, man. Do these musicians know tango? Macumba? Yoruba music? Samba? Bossa nova? Salsa? Cha-cha?
My wife is a very good piano player, but she doesn’t have “it”, she sounds like a robot playing, it’s painful to hear, but she plays in time, all the right notes, but it’s just not there.
In drawing it’s most evident, some people can draw and others can’t, no amount of practice will help me. The emotion side of things kicks in more so with drawing.
Absurdly subjective. If you can neither qualify nor quantify "it" you shouldn't start levying it as some standard of quality.
The differences in timing, attack, etc from a good player and a merely competent player are objectively measurable, but a non expert can hear the difference and form an option without understanding what they are using to discriminate.
Edit: fixed French spelling.
if you're going to pompously bandy a term about, you should at least spell it correctly
also known as "inherently subjective"
just because you deem something to be inexpressible doesn't mean it actually is.
A sample from the young man: https://www.twitch.tv/videos/227075655
Again : just because "you" find that something speaks to "you" is in no way an indication that it possesses some inherent quality. More likely, it's an expression of your personal biases and/or group psychosis.
And you’re comparing them to other random people on Twitch, who might be casual players just noodling around.
So what are we supposed to deduce from this?
I've personally played an instrument since I was 7 and while I was playing every day until I was 18 that didn't automatically make me able to produce good music that gives people goosebumps. But I've met plenty who can, without much practice. In fact their earlier music had an unrefined thing about it that made it unique. They were just good, I guess.
Sorry if that doesn't make any sense.
Generally, I agree that some people clearly have a lot of raw talent.
But also I think we’re very attracted to this idea of the “natural,” who becomes a virtuoso almost effortlessly. In my experience, most people who reach the very highest skill levels have both raw talent and a tremendous work ethic. And a lot of people who show early promise never reach their potential because they won’t put in the time.
Sometimes the distinction between talent and effort isn’t even clear. There’s a documentary called Magnus about the chess champion Magnus Carlsen. It shows him as a kid, thinking about chess from the moment he wakes up to the moment he sleeps (and he probably dreamt about chess too). While others are playing, he’s staring off into space thinking about chess. That is some kind of talent, the ability to maintain complete focus like that over many years.
I’d bet some of these Dota2 players in the top 0.5% spend an extreme amount of time playing or thinking about the game.
Anecdotally, I always use my sister and I as examples. We’re 18 months apart, similar IQ, same upbringing, etc. you can hand her any music producing instrument and she’ll be making music in a few minutes. I am the opposite, barely able to play Mary had a little lamb.
I’m pretty sure you can use tools like Myers Briggs to correlate personality traits to talents in certain areas. Some people are attracted to a career in accounting, which sounds like a vision of hell on earth to others. Ditto for music or other arts.
That isn't remotely true, we just don't teach children to draw. Most teachers believe that it's an innate ability and have no idea how to teach it even if they wanted to. Drawing is a straightforward skill that can be quickly learned, but it's highly counter-intuitive to most people. If you think you can't draw, pick up a copy of Drawing on the Right Side of the Brain by Betty Edwards or Keys to Drawing by Bert Dodson.
I also have no talent for sleight of hand, I’ve never been good at it, until recently, when I deliberately started practivcing for a few hours every single day (learning sleights for card illusions and caradistry). I’m not particularly good yet, but the difference from even a couple of weeks ago is huge.
It takes deliberate practice, motivation and time. Sometimes, you also need a good teacher.
Going by his words: "You don’t have to practice emotion because that comes naturally. Technique is different." By that he means that you do have to practice technique. The title of the article is "No Causal Effect of Music Practice on Ability". That sounds like the words Quincy Jones said directly contradict the words in the title of the article, to me.
In my limited experience, I find the "it" factor which cannot be learned is the ability to keep time; that "swing" you mention. Given that your wife possesses great technical ability, perhaps self-consciousness is a factor in the perceived lack-of-emotion in her playing?
Edit: here he expresses some of the same sentiment https://youtu.be/OG__SwkV3wg
What the study refers to as musical ability is in fact the skill to discern different pitches and rhythms.
Obviously there is an enormous amount af practice required to master an instrument. But I still find it hard to “agree” with the paper, I’ve noticed large differences in how attuned I am to pitch and rhythm depending on how much I’ve practiced. Hearing pitch, at least for me, is a skill that can be both trained and lost. Which makes me think that maybe it is not the total amount of hours practiced that is important but also how recent that practice was. I can find no mention in the paper of how “fresh” the participants skills were. Also, there is a huge difference in how you practice, how efficient it is etc.
To be fair I didn’t fully understand the statistical stuff, but all in all this fells like a somewhat blunt study especially considering the provocative title.
Edit: spelling
It's rather unfortunate that the title is so poor, as the paper itself is good aside from that; the study they mention a few times (Schellenberg & Weiss, 2013) found that "more music practice is significantly associated with better music ability" and I'm glad people are going through the effort of checking results like these for causation vs mere correlation.
Yes, and real musicianship is more complicated still since hearing and playing abilities co-evolve as one cycles round and round a new piece. For example, notes previously played too quietly can be played louder merely by hearing them as louder. At least that's the way I experience it. If true it adds a whole new meaning to 'active listening' since motor activity may be included in the loop.
The paper compares hours of music practice to scores in the Swedish Musical Listening Test (SMLT). It finds no significant correlation, concluding that there is no causal relationship between musical practice and musical skill. The SMLT does not directly measure practical musical skills, but is treated by the authors as a valid proxy measure because it correlates with hours of musical practice (Ullén et al, 2014).
Practice does not improve musical ability because hours of practice are not correlated with scores in the SMLT. The SMLT is a valid proxy measure of musical ability because scores correlate with hours of practice. See the problem?
Ullén, F., Mosing, M. A., Holm, L., Eriksson, H., & Madison, G. (2014). Psychometric properties and heritability of a new online test for musicality, the Swedish Musical Discrimination Test. Personality and Individual Differences, 63, 87–93. (10.1016/j.paid.2014.01.057)
Very interesting article!
Here's an analogy: say you get a discount on your health insurance if you visit a fitness center regularly. Except that going to fitness centers got you excited. You don't get anything out of them. But you do like biking, going on long hikes and you do a lot of physical activity in your work. So just because you don't go to the gym, should you give up on your fitness? Heck no! Lean into what energizes you to stay healthy.
A punitive practice will not force someone's awareness on the practice. Though someone might be bored and put their attention on the practice.
That has been my experience with music growing up and martial arts later. And software programming.
So it translatates in non-fancy-wording. Being able to differenciate rhythms and melodies depends on your genetics and can not be practiced for improvement. But it says nothing about the effect of practicing on being a better music performer.
The claim I'm familiar with is that a performer who practices decreases the probability of a catastrophic mistake in the performance and increases the probability that their musical ideas will be conveyed clearly during the performance. So given two performers or roughly equal ability, hire the one who practices.
On the one hand, I think playing in an ensemble is crucial for learning rhythm discrimination and that there probably would be transfer there. On the other, perfect pitch is all but unlearnable and that presents a huge, entirely genetic advantage for some folks on pitch discrimination tasks.
Woah.
TIL about the Swedish Twin Registry: https://ki.se/en/research/the-swedish-twin-registry
> The SMDT consists of three subtests — pitch, melody, and rhythm discrimination
The test is of fundamental primitives - not about the ability to play a tune. The study doesn't show that practice does not make better music. For example a deaf person who scores 0 on all the tests, can still visually memorize and practice timings and strokes to produce nice songs.
Before I started any sort of training/practice regimen, I'm quite sure I could never identify a tritone or minor third being payed on a piano. After a lot of deliberate ear training, I certainly could. And I suspect most people could as well, but not by just sitting there listening to things you don't understand, but rather by 'deliberate practice' and learning. A major 5th is easily recognized as the interval between the first notes in a popular star wars theme, a minor 3rd is the interval between the first two notes in greensleeves. These sorts of associations were how I taught myself.
This study is baffling...
I'm a lapsed clarinetist. It's been a few years. I wonder if I would have answered "no" to the first question, and been scored as someone with zero hours of practice. Yet I have a couple of thousand lifetime hours of practice.
(But no, most of that practice time did not really drill the skills measured by the SMDT.)
It absolutely is. Practice is necessary but not sufficient for high-level musical performance. The study talks about musical ability in terms of rhythm, melody, and pitch discrimination -- mental skills around music -- and finds that these are not changed with practice. These two statements are not in conflict.
Exactly. I think the paper is trying to establish that there is some kind of inborn talent that is not affected by practice, even if other things (like performing ability) are.
Did they design a test specifically to make them feel special? rhythm is subjective within music (think jazz). A 'good' melody is only subjectively different from a bad one. Maybe pitch discrimination could be genetic; partial deafness.
They may not be technically wrong, but their definition of practice and perfect seem twisted to click bait.
So everybody's lumped together, those that play one instrument vs those than are proficient on 3 or more
Is the difference between Stevie Ray Vaughn and Jimi Hendrix measurable by "rhythm, melody, pitch discrimination" and whatever methodology they used to quantify these variables?
It's like measuring computer scientists ability by how quickly they can implement binary search.
EDIT: it is interesting that musical practice doesn't change those abilities. But I think the title of the article is misleading and grasping.
One the one hand, you have skill, on the other ability. Ability is the potential you have to become good at something, skill is actualized ability. So while I may not be able to e.g. play the piano now (no skill), given enough practice, I can actualize my potential into skill. Potential may place a bound on skill, but the only way to convert potential into skill is through practice.
The Gladwell version of 10,000 hours was quite clearly a 'necessary' interpretation of deliberate practice. You look at success vs. amount practiced and see that those who practice more tend to be more successful, at least to a point. Fairly simple idea, wrapped up in a neat pop-sci quip.
The 'sufficient' argument is a radically different one that suggests that ultimate ability isn't that different between individuals. Immediately, this seems strongly at odds with other observed intellectual metrics such as IQ which is very highly heritable (h = 0.7 to 0.8) and does not suffer from many of the data quality issues that other metrics do.
A much better title for this paper would end with 'ultimate ability' or 'potential' or something, because practice is obviously necessary. There's just a lot of people talking past each other here, and several that clearly didn't read even the introduction to the paper which discusses this difference.
Learning to play an instrument or sing is not the same as studying pitch differentiation or rhythm. There are a surprising number of successful musicians who are bad at these things. But the method laid out here is utterly meaningless. You might assume that practicing the violin would have some affect on these things, but it doesn't necessarily. Many people simply learn these things by wrote for specific pieces of music from their guidance in private lessons and teachers.
You might also assume that someone who practices a lot and got a degree is also good at Music Theory or Music History, but that's also false. There are tons of professional classical musicians with graduate and Doctorate degrees who are bad at all the things measured in this study as well as history and theory.
I'm really shocked that the authors couldn't figure out that they were measuring two completely different things and calling them the same thing. It almost seems intentional for the sake of publishing a controversial result.
People absolutely get better at these specific tasks when they practice _these specific tasks_. Every music curriculum in world specifically requires classes in these exact tasks, although the bar for minimal competence is extremely low.
I run a small side business tutoring people who are already professional classical musicians who want to get into a DMA program. They have to demonstrate competency at these tasks, as well as some basic knowledge of Theory and History. They can't; even though they are top-level performers on their instrument and practice hours every day.
I offer a 12-week crash course that covers these types of tasks and as well as the knowledge component, and people usually find me because they've already applied to a DMA program and been rejected for failing these prerequisites. I know for a fact that they get better with deliberate practice because after working with a couple of dozen clients, every single one of them gets in on the second try.
Not to mention the fact that every single day, there are students in Music degrees who start off really sucking at these specific measures who get enough better at them to pass the class at the end of the semester, through very specifically designed exercises that target these abilities.
This whole study is an infuriating waste of time and money, and it's an example of some really bad science that's floating around there in the world I really have a hard time believing this was accepted anywhere for publication.
maybe if you read my paper you would see how you don't need any caloric intake to survive. (If I redefine all the words I just used, in the body of my paper.) Hence my "why the obviously false titles."
There's nothing "obvious", unless you can counter both the results and the explanations given in the study. And they haven't redefined anything.