back
39 comments
This seems like a very muddled claim to me. I won't refute that there's some kernel of truth in there somewhere, but using anecdotal evidence to loosely tie one's own theory together rarely ends well, especially when one's claim would be beneficial to oneself if true.

A perhaps more plausible and/or stronger explanation for programmers' disproportional display of music talent as a group:

To gain the experience necessary to label oneself a programmer, one has likely a >lot< of time on a computer. Whether that computer exists at home or at a school/university, we can already likely assume (especially for the population of "professional" programmers, due to the lesser availability of this necessary resource in past decades and the length of time necessary to become a "professional") that the group we're talking about trends toward the higher income side of the population. That alone would make the group as a whole far more likely to have ever picked up a musical instrument in the first place, let alone to have received the countless hours of necessary instruction.

Additionally, I think we can all agree programmers are disproprotionally likely to be homebodies. If one spends a lot of time at home, one has more opportunities for practice.

Lastly, if one has the patience and determination necessary to learn enough programming to label oneself a programmer (especially a "professional"), then one is more likely than the rest of the population to have the same patience and determination necessary to learn another skill which requires countless hours of disciplined study-- like music.

It goes deeper than amount of time spent. There are differences in the MRI scans of people who have spent time doing substantive work in music before the age of 12, and those who started much later. I've also taught music, and let me tell you, the difference is stark! It's much the same as learning a spoken language. There is a big difference between 10,000 hours spent learning spoken languages during the K-12 years, and 10,000 hours in your late 30's!

Teaching adults who had no exposure to making music as a kid is like trying to teach someone who can't see color to paint a sunset! There are some adults who can't keep rhythm with any meaningful accuracy, and can't hum a tune back at you. There is some entire facility which is missing. I've had similar experiences teaching programming.

I've recently wondered, is the common knowledge that kids pick up languages quickly and better than adults really true?

Adults have a much wider range of skills and far more knowledge than children, adults already have some of the most important aspects of thinking down: a set of concepts about the world. Yet many struggle to learn a second language past a certain time in their life.

Clearly children learn languages better than adults, a near 100% success rate for children picking up their first language, to what may be something less than 50% success rate for adults (guessing these numbers, but its clearly lower for adults).

As a child though, what do you spend nearly every waking minute doing? practicing and learning language, usually from their mother. Where as adults, usually they give up not because its impossible, but because its too hard -- and they have better things to do!

Is ten thousand hours of language study as a kid really worth ten thousand hours of language study as an adult?

I would be interested to hear if anyone knows of research looking into the actual hours/work put in to study a second language to get to some certain level of proficiency where the subjects started learning at different ages (5 years, 10 years, 15, 30 years old etc.)

I've recently wondered, is the common knowledge that kids pick up languages quickly and better than adults really true?

The question is insufficiently precisely formulated to be answered. Unreferenced BS follows.

As you pointed out children don't actually pick uplanguages that fast. It takes a motivated English native speaker about 600 hours to achieve near native fluency in German if they have any gift for languages. That's just classroom and home study, not talking immersion. It takes on the order of 10,000 hours for children to learn a language with native fluency.

Even for children who started learning a language by immersion before 14 (approximate cut-off point for native fluency) there's a decent statistical relationship between vocabulary and the age at which they started learning. Obviously if you moved with your parents to X at ten and are a voracious reader you'll beat the deprived kid from the slum estate, statistical relationship.

Learning a language that's closely related to one you already speak isn't that hard if you are totally immersed. Grammar books, vocabulary lists, dictionaries, helpful native speakers all help tremendously. It is possible but incredibly hard for an 'adult' (14+) to learn a foreign language natively, unless they're a biological freak (hyperpolyglot, remembered from SciAm, treat with appropriate caution)

Learning a language that's closely related to one you already speak isn't that hard if you are totally immersed. Grammar books, vocabulary lists, dictionaries, helpful native speakers all help tremendously.

Yes, being quite fluent in a language is just a matter of effort. But really speaking like a native -- this is exceptionally hard. My dad's English is incredibly good for someone born in Korea. His grammar is impeccable, and he's completely fluent. But I still catch that he still has to think a little bit about L and R sometimes. (And he's been at it for 60 years!)

A measurable difference in brain scans does not by itself amount to enough evidence to claim that programming and musical ability are intrinsically linked.

All I'm saying is two things:

a) Extraordinary claims require extraordinary evidence (note: neither a vague, nice-sounding theory that everyone would like to believe to be true nor anecdotal evidence qualify) and b) I feel an application of Occam's Razor is in order.

There is a very important scene in "Tous les Matins du Monde" where Gerard Depardieu is pretending to play a Viola de Gamba. (Ancestor of the Cello) It's the climactic scene, actually.

This scene works very well for most people who have never played an instrument. Gerard Depardieu is considered a good actor by many, and he's doing all sorts of things right as far as that goes. However, people who have played an instrument often cringe when watching that scene, because it's glaringly obvious that Mr. Depardieu is not playing the instrument. He's faking it so badly, it makes many musicians consider the scene hoky and disappointing. (It's still a great movie, though.)

What's really interesting, is that musicians who know nothing about playing a Viola de Gamba can still tell. I play wind instruments, and my fingers are doing nothing that resembles what a bowed string player has to do. There is no resemblance at all -- some things are diametrically opposed, and while I do nothing like bowing, I also need to incorporate breathing to play, in a way which is not at all needed for playing a string. Yet I can still tell immediately. This difference is very stark, and you'd probably see much better discrimination in a double-blind test than with 128 bit MP3's versus 320 bits. Note that there is a well known mathematical and physiological basis for the latter.

Yeah, I play strings and I hated that scene. You can even deduce from this scene that Depardieu hasn't learned to play any type of instrument, which can be said for a lot of Hollywood actors faking it on camera.

Although, there is the rare breed of actor that plays an instrument.

Towle_,

Thank you for the followup. I'm not trying to do science here, just understand my own experience.

Regarding the correlation between musicians and programmers being the result of socioeconomic status, that doesn't match my experience. I had lots of high school friends from similar backgrounds, and very few of them became musicians.

My experience is that the brain "peculiarities" that dispose one for accomplishment in programming are the same structures that dispose one for accomplishment in music. I'd love to see what a scientist would make of this claim.

Music also goes deeper than pattern matching. Often, it is a game of awareness! A really good musician also becomes aware of complex secondary effects, both in the mix of sounds produced and the effect those sounds have on the emotions and "kinesthetic sympathies" of listeners. (By "kinestheic sympathies" I mean a gut reaction people can have to the physics of movement implied by asymmetries in rhythm. "It don't mean a thing, if it ain't got that swing!")

Awesome acoustic musicians are aware of little nuances, and can exploit them to great effect, much as a great chef can bring out a little edge of nuttiness or caramel in a dish that enhances the whole. Sometimes these effects can be stretched out over time, and at 2 or three levels of granularity simultaneously. (The tasty thing you are doing this time through, is also setting up a resonance with something you are doing the next time, which is part of an overall structure of drama and anticipation you're setting up over the whole song.) I've started delving into pop influenced by electronica, and find there are many of these exquisite things in electronic music as well.

To bring this back to programming: there is also a game of awareness that can be played while programming. There are may levels of granularity, and many parts coming together in ways one might not yet be aware of. Often, it's bringing such things into awareness that results in that great refactoring.

What you are calling "awareness" sounds like what I mean by pattern matching. When I'm "on" as a musician, this "pattern matching" operates on a variety of time scales--multiple concerts, this concert, this piece, this verse, this phrase, this chord-- and multiple people scales--me, my fellow musicians, different people in the audience, the audience as a whole.

At the very highest level (perhaps five times in my life) any consciousness of all this disappears, even though I'm sure all those factors are still influencing my behavior. Strangely, this is absolutely sublime in music but I tend to produce code I later regret when programming in this zone.

"What you are calling "awareness" sounds like what I mean by pattern matching"

Perhaps pattern matching is involved. What I'm thinking of are occasions when one is listening to a very good musician and one is very taken by a particular phrase or lick. The pleasure in that moment is intense, but it's often hard to put the cause of that pleasure into words or even to replicate it musically. I often find that I have to go back and analyze what happened, sometimes by playing back a recording at half speed. Even then, I can fail to replicate the essence of what was good in that moment. There's a certain discipline of awareness one has to practice. Even though I'm able to produce a resemblance of what I've heard, this often lacks the essence of what gave me the pleasure. I have to be aware that I'm reproducing the right thing sonically, and not just taking pleasure from mastering a new sequence of movements. (And it's quite common to meet musicians who just stop at doing the latter.)

Pattern matching might be involved, but it's only part of it. There's a kind of introspection involved in figuring out what made it work. I have to think about how I'm feeling and connect that causally to some particular nuance or combination of things.

This emotional resonance is still pattern matching, but of a different kind. Rather than match patterns between notes you're matching patterns between note patterns and the emotions they created. The awareness you're talking about, in my view, is way of emotionally visualizing the reaction. In a similar way we try to visualize something before actually building it.
The awareness you're talking about, in my view, is way of emotionally visualizing the reaction. In a similar way we try to visualize something before actually building it.

That is emphatically not it! Well, maybe it's involved but only in the latter part. I'm not just talking about conceptualization. I'm talking about getting aware that there's something to conceptualize in the first place. Pattern matching is involved, but I'm not sure there's all there is to the sort of introspection I'm thinking about.

Calling yourself a programmer and calling yourself a musician are both fairly easy. I would say a lot of programmers are musicians because a lot of people are musicians. I know a lot of musicians who are not programmers.

I think that if you actually look at the most talented musicians in the world, they are probably not programmers either. The amount of dedication it gets to reach the top makes it difficult to do both.

That's certainly a possibility, that programming accomplishment and musical accomplishment don't correlate at all and it's just sampling error. However, I experience both and I was trying to identify the commonality, in answer to a question that has been discussed ad infinitum. It would be interesting to me if the question applied to people in general. At the moment I'm glad to have made my world slightly more understandable.
I sing, play guitar, bass, and saxophone, as well as enough drums and keyboards to get by in rock or pop settings.

With my current group we've started to use a heavy dosage of Max/MSP for audio and Quartz Composer for visuals... my professional programming experience really helps with creating really good interfaces for the software (not just UI, programmatic interfaces as well), so extending or modifying the software happens at a pace that keeps up with the rest of the band... and being that music is prety much realtime...

If you're interested, we've got some demos up... http://redblueyellow.com

Of course I did our website as well... Rails ;)

And how many bands have a github account? Sorry, no public repos yet... http://github.com/redblueyellow

At one point we had three software developers in the band, but we've slimmed down due to gasp, guys getting jobs in the industry elsewhere (other than SF, I know, right?) and now I'm the lone coder.

I'd say I'm about 50/50 when it comes to music... I've had periods where I did it full-time, and I've had periods where I've coded full-time. Right now I'm burning the candle at both ends with a contract job and spending at least 30 hours a week on the band.

I know quite a few professional programmers who are also professional or semi-pro/very dedicated musicians.

I think there is more to it than mere correlation.

And can I say that the 10,000 hour thing is WAY WAY off?

I'll put it in musical terms:

I started playing the saxophone in 5th grade. I played for 2 years. I switched to guitar. About 4 years ago I picked up a sax because the band I was in needed some horns. After about a week, I was 100 times better than I ever was before. So what happened? Well, I learned more about music in the mean time. I'm sure I could pick up any instrument much, much quicker than someone who has never played any instruments. What am I getting at? Well, a lot of things in life have more than just a passing similarity. They have core fundamentals that are shared. I am completely convinced that this applies to music and software development, although I have no idea why.

Humans are pattern matching machines. I think it's the root of our intelligence - it's certainly necessary for our survival. Basic pattern matching is required to, say, determine what kinds of plants are safe to eat, and to figure out where big game is and to predict where it will be.
What I noticed is that 1) some people are better at matching patterns than others, 2) there are different kinds of patterns (social, symbolic, spatial, etc.), and 3) some people have a particularly effective reinforcement loop around pattern matching.
Have you read on intelligence? Jeff Hawkins, talks about how everything we do is about detecting whats different in the patterns.
No, I haven't. I heard an interview with him, though, that makes me think we're thinking about similar ideas.
Learning to play a musical instrument changes the brain.

http://www.articlesbase.com/art-and-entertainment-articles/b...

Isn't it accurate to say learning anything changes the brain, since that's what leaning is?
I agree and since the link doesn't seem to work for me, I can't comment on the article.

Perhaps learning music changes the brain in a way that is different than say, learning how to cook French food or to fix a computer?

Actually, if you are a excellent French cook, I suspect that your brain was changed in much the same way. Fixing a computer, not so much. See my post about the "game of awareness" in music.
Yes, but it's well known that certain kinds of changes are qualitatively different. There is a big difference between learning how to speak French when you are a child, versus someone learning in middle age. There are also big quantitative differences in MRI scans of people listening to music between those who learned how to play an instrument before the age of 12 versus those with no such background.
Since I've learned to play guitar (and studied music theory) my mind has opened up to the possibilities of learning and I feel like a smarter and sharper person. I can't say with any scientific evidence how music has changed my mind but I'm going to agree with that general statement.

Additionally, I find that when programming a short musical break will rejuvenate my mind and get me right back into the flow of code.

I find I code more quickly and elegantly when I'm listening to music - especially ambient music without a lot of sharp edges that throw me out of flow state.
I'd be curious to see a poll of HN readers asking how many enjoy/excel at both programming and music.

I'll never win any awards for either my code or my compositions, but I love programming and I love playing music (rhythm guitar and drums, mainly).

Unfortunately for the thesis, we have a well-established counterexample.

http://www.jwz.org/hacks/why-cooperation-with-rms-is-impossi...

Regardless his singing abilities, that's a beautiful song IMO. He seems to have taste at least.

http://en.wikipedia.org/wiki/Sadi_Moma

Awful singer, good dancer: http://www.youtube.com/watch?v=Pube5Aynsls
The good news about psychology is that a counterexample does not invalidate the thesis the way it does in math.
We probably disagree on whether this is, in fact, 'good news'.
What is the basis for claiming programmers are musically talented? I know a number of very talented musicians and a number of very talented programmers, but the overlap between the two is pretty close to zero. Sure I know programmers who can string together a few chords and musicians who can hack together a website in php, but I've never seen anything to make me think that the overlap between programming and playing a musical instrument is any larger or more significant than the overlap between any other profession and playing a musical instrument.
AFAIK the correlation is anecdotal and no more. However, I've had enough bar discussions of "why do musical and programming accomplishment go together" that I was glad to find an answer that made sense to me. I'd love to see real research to confirm or deny the correlation.
This claim is just a subset of people-who-excel-at-anything-are-able-to-relate-to-excellence-at-anything-else
I've certainly seen that kind of argument. What makes me think I haven't fallen into that fallacy (although it is a possibility) is that I freely admit that I'm terrible at lots of things (like reading people) and that I'm good at other things (like baking or cheesemaking) that have nothing in common with my experience of programming and music.