The 93% mentioned is far beyond the expectations of most. Prior to the 20th century (at least in the upper class) a child was taught every thing from Greek and Latin to water color painting without regard to 'talent'. Examine the life of Ada Augusta. While there is no doubt that she was talented, her studies were not predicated on that talent. Today most people foreclose on what they might become accomplished at because they say things like 'I can't draw a straight line' or 'whistle a note' or similar such. Again they have no understanding of just how good that 93% level is. I can't teach you to engrave as well as Rembrandt, but I can teach you to engrave as well as the majority of professionals.
"Talent, Practice, Scientists, and Journalists: Can't We All Just Get Along?" by Scott Barry Kaufman, co-editor of the Cambridge Handbook of Intelligence,
The article and the conclusions of the underlying study bring up more questions than answers for me. Such as:
1) Does the importance of talent vary by domain?
2) They show working memory matters. Does working memory matter by different amounts in different domains?
3) What other "talent" factors work across different domains?
4) To what extent does having natural talent in an area lead to the development of poor work habits (because things come so easily) and thus long-term underperformance?
I don't have answers, but I believe that drawing firm conclusions about the talent/practice question is premature.
WRT 2: Working memory capacity is (according to one well-regarded theoretical model) actually a series of different components - the phonological loop (stores verbal/phonological/textual data, can hold just about the seven digits of a phone number), the visual/spatial sketchpad (can hold representations of images - not sure there's as straightforward of a measure of its capacity), the episodic buffer (not so clear it exists - separate component for integrating different components), and the executive system (controls attention, attending to specific bits of input). Many of these are confirmed theoretically by constructing experiments to overload one memory component, and then checking to see what other capabilities still have spare capacity.
While an individual's capacity in each of these components is pretty fixed, there is some variance in what component of working memory a task exercises.
In addition it might be that some tasks might lend themselves to good compression. For example, an even lightly experienced programmer thinking about a for loop might not require one "slot" of phonological memory each for the index, start, finish, and label, but could rather represent the entire structure as "for loop" in their phonological memory, this single unit acting as a reference to the long-term memory representation of iteration. In this way, while working memory in its raw form is rather constant over different types of tasks, the actual amount of semantic information it contains may vary widely with task domain.
Lastly, mnemonic strategies matter a lot. For example, it might make a big difference to a music learner whether, when sight-reading, they process a piece in their phonlogical loop ("the notes are A-G-G-A") or in their visual-spatial sketchpad ("This bunch of dots is the shape of the tune on the page").
A primary question, of course, is how malleable working memory is to practice effects. The experimental research shows that improvement in domain-specific working memory is very malleable under deliberate practice.
Does skill at piano sight reading generalize to other skills?
My experience is that a greater working memory obviously helps with every mental task. I constantly run into the limits of my working memory; problems too complex to grasp.
If I immerse myself in the problem for a while (e.g. 3 days), it often becomes clear. I think that some aspects transfer into long-term memory, until what remains fits into short-term memory. Another deeper effect is when I see a new way of looking at the problem that makes it simpler; a way to divide it into parts, or layers, or aspects that can be dealt with independently.
It's frustrating to have limited working memory, but I console myself with the fundamental truth that everyone has limited working memory. A far more important ability is being able to handle problems that exceed one's working memory - with a few steps, you easily exceed the working memory of anyone.
Another benefit is that this division into parts/layers/aspects is a theory of the problem, which can be transfered to others. By finding a simpler, better way to think about this problem, you can save everyone time and effort (including super-geniuses, because it frees up their working memory). Even wrap it up in a product and sell it, which saves even more. And even if a super-genius could effortlessly grasp it in working memory, he/she would have no interest in finding a simplifying theory of something already trivial, would feel no value in it if they did see it, and couldn't believe there's be any value in bringing a "solution to a non-problem" to others. It takes someone like me. And of course, the real super-geniuses end up working at their limits, same as me.
Of course, it would always be nice to have just a little bit more working memory. One reason to get enough sleep, food, exercise.
I don't think genius works that way. I am may not be a 'genius' but I am close enough to get the occasional taste of it. First off you have more working memory, and you also think faster, but that enables you to think deeper. Which let's you find short cuts sooner, and those shortcuts let you think about it even deeper. What's shocking is just how fast this can be.
EX: As a sophomore in collage I took an AI course and the teacher was introducing the idea of using heuristics optimize searching for the ideal path. He basically said find the best node, explore paths from there then resort the node list. And I was thinking how useful that could be but also that you don't really need to completely sort the list just keep buckets of nodes and only sort the best bucket. And then I started running with that idea thinking of various was of optimizing it for multiple CPU's etc. So the teacher and a few minutes into the lecture the teacher asked "does everyone understand why you need to resort the list" and I was said something like "Sure, but I don't think you need to resort the list just find the new best node(s)". Which started a discussion that just confused most of the students.
Thinking back I can see how that was not really the right thing to say at the time. But, I can only imagine someone that's profoundly gifted and has that type of thinking happen not just from time to time but most of the time. Speech literally becomes to slow to have a meaningful conversation with most people. And you start refining your ideas in the middle of a sentience as you consider varies new concepts.
PS: As to the value of practice it's IMO a question of how important that last 5-10% is. Get to the top 70% in your field and you can can be a vary well paid doctor, accountant, engineer etc, but being a physicist is judged on a different scale. A pianist or football player that's 5% worse than the best in their field is probably better of looking for something else to do.
Let's say that there are 100s of unique 'traits' that correlate with talent at a complicated skill, like piano playing. And, the authors have correlated one of those traits, working memory, with better piano playing.
That's great. Does that mean I will be a great pianist if I have awesome working memory? No - I still need to practice and lots of other traits that make up that talent.
I will hypothesize that every complicated skill requires 100s of unique traits to perform that skill well. I will also hypothesize that each of us Humans have 1000s or 10000s of unique traits, and for each of these we fit onto some Normal curve.
Does that mean that we wait around to discover that we are two standard deviations from the mean on every trait that is required to accomplish a complex skill? Hell no. We try to identify the few or several traits for which we excel for that task, we emphasize those traits, and we practice.
Yes - there's no question that we are all made differently. Does that matter? It's hard to say, until we try.
Things like aptitude tests seem to test for a very narrow range of traits. And, all they do is discourage those that score poorly from pursuing skills, even if those who score poorly may have other traits that more than make up the difference.
Isn't fluid intelligence variable and hence working memory? Didn't they say that dual n-back can increase intelligence?
I wish they had an equivalent of dual n-back for motivation. I think I have always said it, people on HN don't need more intelligence; they more need to actually get things done.
The problem with motivation and anything technique related to it is you have to be motivated to execute said technique.
Poor working memory is a motivation problem for some people who find it too difficult to do what they want to do.
Other people need to hack their principles, their health, their environment, etc. to do what they want. I believe all of those have a "gamified" way to increase them the way dual-n-back improves working memory, though they have not all been developed yet.
In all seriousness, motivation is pretty heavily dopamine mediated. You can train for willpower, but it takes a long time and a lot of effort in return for modest gains.
Very true. The result of this for people making strategic decisions regarding their lives is that if the field puts high value on positional factors, you should only engage in practice in that field if you have natural talent. For example, being in the 93rd percentile of basketball players (or startup founders) isn't going to make you millions of dollars (The value is in being better with respect to others, not with being good in absolute terms. Only one person gets to be the best.).
But being in the 93rd percent of public equity investors will make you mightily rich.
If you don't practice, it doesn't matter how much talent you have.
Exactly, greats don't just show up great. Carmack is considered great in the field of graphics programming, and if the Masters of Doom book is correct he is a mix of tons of talent and lots of practice.
Very true. For some pretty far is good enough and it's where they're happy and want to be. However there are others who do strive for #1. I'm not talking about just generally as in how everyone wants to be the best but instead what I mean are those special few who need to take over the world and are meant to. Those people need that extra 7% but most of us don't.
I for one have always known that I am, in an irreducible way, smarter and indeed _better_ than everyone around me. I am gratified to see this empirical reality reflected in scientific literature, and see no reason to question these findings.
I think the problem is the quality of the deliberation of "deliberate practice." There are metaskills, such as reason, logic, understanding of cause and effect, statistical analysis, etc. that effect the acquisition of all skills, partially through the method of improving the efficiency of memory. If you learn one skill that has five objects, consisting of two transformations of two of those objects into another two of those objects, and three relationships between those two resultant objects that result in the fifth object, you've just learned a complicated skill. But if you've learned how to find potential isomorphisms between those objects, transformations, and relationships and other objects, transformations, and relationships in the world through the use of reason, logic, understanding of cause and effect, and statistics, and how to vet them well, then each further discovery will immediately generate skills that weren't possessed before, and some of them may even be novel. Without the ability to recognize those isomorphisms, each skill would have to be learned separately, and remembered separately. It's like the difference between languages like Old English, where to pluralize a word or to create a possessive would require learning another word, and normalized languages like most of Modern English or the Romance languages where you can just apply a transformation to words that you already know. In the leap from Old English to Modern English, English speakers were given a gift of an expanded memory.
I hope that didn't sound like word salad.
Basically, if I had to get by on sheer memory, I'd be screwed, because mine is miserable. What I can do is reason to the same place under the same conditions often enough that my memory will eventually leap there out of habit. To get good results out of this requires a strong foundation that sadly both isn't formally taught or emphasized in our educational system (outside of mathematics and the hard sciences), just gleaned through the sheer effort of having to memorize an enormous amount of information. I don't deny that some genetically predetermined concentrations of proteins in the brain could give you a leg up on this, but I deny that any brain that has all of its pieces couldn't possibly outperform every other brain on this planet in any particular area.
I operate under the assumption that advances in reason will be made and distributed during millennia to come that will give the average 10 year old the equipment to understand within a day things that would have taken a person in the early 21st century a dozen lifetimes of what counts as "deliberate practice" now.
We invented language to spread information and ways to reason about that information, and it's worked. First we spoke it, then we wrote it, then we manufactured it, now we're almost bathing in it. If we manage not to 'splode ourselves within the next 100,000 years, we should be as advanced in reasoning over our current selves as our current selves are over who we were 20,000 years ago. Why 100,000? I'm assuming diminishing returns because of the discoveries that we sometimes make of walls like quantum physics, where we discover things that are even theoretically unknowable. Probably a primitive assumption:)
At this point in time, we jump to mystical theories of the undetectable secret superiority of the successful, after defining superiority as being successful (at the rate of gaining returns with practice), which is completely circular logic that discounts "deliberate practice" as a skill in itself.
The author shows that actual numbers bear this out.
</rant>
He's got a lot of other interesting articles about the dynamics of deliberate practice. Thanks for the link.
Talent is very important. Unfortunately, when you start saying that mere mortals not born with some innate talent for X task are destined to forever be not as good and less successful everyone throws a fit because the reality is that the majority of us just don't have that talent we crave. It implies that those without the talent won't get ahead and that's really upsetting.
I'm still on the fence as to whether to lean on the side this post takes or the original study that it's arguing against but I think this whole discussion has a lot to do with something else entirely. Please try to follow my logic here as I think it's am interesting connection.
The American Dream. The story we're told about America being a meritocracy and if we just work hard enough we'll all make it. This myth is pounded into all US born citizens from the time we're first enrolled in school and maybe earlier. But unfortunately this isn't true. In America, if you work really hard you'll get whatever crumbs your masters decide to give you. Power is inherited and not earned more often than not and it's rare that someone breaks out of the ranks and really shakes up the system.
So when you imply that talent counts more than hard work it shakes you to your core because of the popular belief that you can do anything if you just try hard enough. But what if we're wrong to refute this? What if talent really does matter more? Acknowledging something like this would surely do more to help you overcome that hurdle than trying to disprove it would. It's like the American Dream. Do I bitch about how it isn't fair and try to convince myself it's real or do I accept it and use my newfound understanding to find away around the problem rather than trying to convince myself it doesn't exist.
Again, I'm on the fence here but if I were to play devil's advocate and assume talent matters and practice doesn't, wouldn't that line of thought be more helpful?
Based entirely on practical experiences (both learning and teaching several subjects), talent does matter at two points:
When you just entered a new field, talent will propel you ahead faster, for a while. (I'd say it's about 5%-10% of the way to mastery, but that's a wild guess).
Then, both talented and untalented people need to work equally hard. Maybe the talented ones get ahead slightly faster, but overall it's a level playing field.
Then, once you approach mastery (let's say at 93%, for arguments sake), talent becomes important again. Without it, you can work as hard as you want, you will not progress, or you will progress very, very slowly. But even if you are talented, that last step takes _insane_ dedication.
I'm not talking about the "mastery" level in the 10,000 hour sense. I'm talking about being at the absolute top of your field. Talent determines the upper limit.
Which means for most people, having talent or not only matters at the beginning of the journey. (Because few strive to attain absolute mastery)
Edit: And I just realized that I could've skipped half of my post by just reading the article :)
