16 March 2009

Hangin' at the Love Sculpture

There's a reproduction of the famous Philadelphia LOVE sculpture near Shinjuku. It's a popular meeting place, not least because of the unusual circular traffic light fixture in front of the sculpture. This couple ambles past on a chilly night.

From MyokoSuginohara-Mar2009

The Japanese Face of Abandonment

It took me quite a while to tune into it, but there are a lot of abandoned building in Tokyo.

The Japanese abandoned building movement was pioneered by the irregularly-published magazine Land's End, which I was lucky enough to meet the publishers of at Design Festa. But, once I started watching closely, there are abandoned buildings all over Tokyo. There's a huge former corporate reasearch center sitting empty right near Hiro-o station, and closer to home, there's an abandoned junior high school just down the street from my house:

It's very Japanese and all: the place isn't overflowing with garbage or anything. In fact, that's what makes abandoned buildings relatively hard to spot in Tokyo: they look a lot like any other building, except no one goes in or out.

The sign says, basically, "Keep Out."

It's not hard to figure out what's going on with this particular building: the Japanese birthrate is incredibly low (only Italy is lower among developed countries), and so there's just no need for a lot of former schools. And this is in suburban Tokyo, which is doing relatively well in terms of number of children: smaller and more rural areas are really lacking in kids. So, at some point, they just decided to lock the place up and hope more kids show up someday. It'll probably be this way for 20+ years.

15 March 2009

You deserve a hot dog today... and it's what's for breakfast!

McDonald's Japan has a new ad campaign here. They really don't impact on my life much (I'm not a big fast food eater), so the last campaign (the introduction of the Quarter Pounder to Japan) didn't really catch my eye. But this one is too good to ignore: they're introduced the McDonald's hot dog:

McDonald's hot dog

Better yet, it's (wait for it)... a breakfast food! That's right, the hot dog isn't on the normal menu, it's on the breakfast menu, available only until 10:30am. So the next time you wake up in the morning, run on over to MickeyD's for a hot dog! Who needs cereal anyway.

13 March 2009

Malcolm Gladwell's "Outliers"

This started as reactions to my Facebook status, but my response got a little long, so I moved it here. In an attempt to, as Dan said, categorize my discontent, here are my ripostes to several sections of the book.

The Hockey Example

In the initial hockey-player chapter, Gladwell points out a huge statistical anomaly associated with top-end sport figures in many sports and countries: their birthdays are not randomly scattered throughout the year, but rather concentrated in a 3-4 month period. His theory is that when elite teams are first formed (for the sports and countries in question, around the age of 10), the fact that the kids born just after the eligibility cutoff date are older and therefore bigger gives them an advantage. This initial advantage is maintained all the way through the school experience (since it's judged by cutoff date) and is strongly reinforced by the extra training, coaching, and practice received by the athletes chosen for those initial elite teams.

This all seems pretty reasonable, and he argues his case well; as he says, kids born in the second half of the eligibility period have a much lower chance of getting into elite teams in sports and countries where those teams form early, such as Canada's hockey leagues.

But... so? At one point Gladwell states, "If we took coutermeasures such as running hockey leagues twice a year, Canada would have twice as many top-level hockey players." That statement is a stunning example of what he later calls lack of practical intelligence!

No, we wouldn't have twice as many top-level hockey players; the number of such players is by definition constant. It's limited by the number of elite junior hockey teams, the number of elite junior hockey coaches, etc. There are only be one champion, no matter how many compete. So, running a second set of hockey leagues (which would lead to playing hockey all summer, by the way) wouldn't affect Canada's supply of hockey stars at all - it would, logically, raise the level of skill of those stars slightly.

But how much would it even do that? As Gladwell points out, the initial difference among players is very slight; most of the reason hockey stars are better than lower-level players is the additional coaching, support, and practice time. Since those effects wouldn't change, the huge overhead of adding a second season would only address inequity within that small initial difference. In practice, the overhead of running the alternate leagues would probably dilute the effectiveness of coaches and training, and the net result might well be a lowering of the level of play at the end of the program.

Gladwell states that hockey is not a meritocracy because of its prejudice towards people born in the first part of the year. Actually, all meritocracies have rules, and people are advanced on the basis of those rules. For example, the Chinese mandarinate favored those whose minds were well-suited to rote memorization, and systematically discriminated against those who practiced innovative thinking. That may have been bad for the country in the long run, but it was still a widely recognized meritocracy.

Gladwell is correct that the rules for Canadian hockey contains clauses that favor those born in the first few months of the year. That doesn't make it invalid, or not a meritocracy.

The Self-made Man Story

Gladwell has a strong aversion to the story of an individual succeeding on their own merits. In the first chapter, Gladwell baldly declares that "In 'Outliers', I want to convince you that these kinds of personal explanations of success don't work." Instead, he shows how personal computer entrepreneurs of the late 1970s, American industrialists of the 1830s, his own grandmother, and New York Jewish lawyers of the 1930s were all born at the right time and place to have the best chance of success in their chosen field. He writes, "What truly distinguishes their histories is not their extraordinary talent but their extraordinary opportunities."

I read through his examples and stories and come away with the utterly opposite conclusion. To me, what distinguishes all of the people he examines, very notably his maternal grandmother, is their individual actions and drive, and the way they took advantage of their opportunities. The individuals he studies were demographically advantaged for various reasons; there were accidents of timing and history that gave them opportunities not available to their peers just a few years in either direction. In fact, virtually every person he studies is quite aware they were lucky for their opportunities and says so.

But does that means what distinguishes them is their luck? I have been around elite schools for a fair portion of my life, I've come to realize that a lot of people have pretty amazing, lucky opportunities. Few of us take advantage of those to achieve stunning levels of success (and many fail to even achieve moderate levels of happiness). While those of us who have worked around Steve Jobs or Bill Joy may have either liked or disliked the experience, few will assert that these are ordinary individuals, even if ranked among those with the opportunity to attend elite schools.

In the end, it probably comes down to the question, "distinguishes them from who?" If you want to ask what distinguishes Robert Oppenheimer from a peasant tenant farmer in Laxmangarh village in India, you would have say it was that Oppenheimer was born into a middle-class family, and that he was smart. If you want to ask what distinguishes Robert Oppenheimer from the tens of thousands of other smart, middle- or upper-class men who graduated from college the same year he did, Gladwell's explanation no longer suffices.

The Practical Intelligence behind the Self-Made Man Story

An even bigger thing I think Gladwell misses is the question of why we are fascinated by the "Horatio Alger" stories that surround many of these people (or to use his phrasing, why we focus on what these people are like, rather than focusing on the time, socioeconomic class, and place of their birth). The reason that we humans focus on those aspects of the stories is simple: we can't change our birth year, birth country or native culture. We can only change our future behavior.

Knowing that to be a billionaire Silicon Valley success story I need to have been born five years earlier than I was is of no use to me in affecting my marginal future success. Knowing that had I been born in China I would likely have learned to count earlier in my childhood is of no use to me in affecting my marginal future success. Etc., etc. I believe the reason we're more interested in stories about successful people's personal habits, business strategies, and so forth is fundamentally because those elements have potential applicability to our own lives; knowing that their junior high school had access to free timesharing terminals, or that they were born in a certain year, has no applicability to our future choices.

On the other hand, knowing that in his senior year of high school, when Bill Gates got an opportunity to go onsite for three months to write commercial software at Bonneville Power Station, that he grabbed it and managed to talk his high school administration into classifying this as "independent study," is valuable. It provides a lesson in taking advantage of opportunities even if they seem impossible.

The Unasked Question about the 10,000-hour Rule

One of the sections of the book I liked somewhat better is the section about the 10,000-hour rule: the idea that true mastery of a skill requires 10,000 hours spent honing that skill. I haven't specifically run into this number before, but it certainly resonates with me.

However, this chapter, especially the section about music students at Berlin's Academy of Music, leaves the most important question unanswered. For review, the basic finding was that the teacher's evaluation of the student's talent level correlated nearly perfectly with the total number of hours the students had spent practicing their instrument: students evaluated the highest had spent 10,000 hours; students in the next stratum 8,000 hours; and the lower-ranked students 4,000 hours.

Gladwell then makes the classic statistics mistake of equating correlation with causality: he assumes that the reason the top-ranked students are top-ranked is because of their playing time. In fact, he never asks why those students got the most playing time.

In fact, as anyone who's played an instrument and especially anyone who's been forced to take music lessons they didn't enjoy can attest, a very plausible reason is that when you're good at something, it is usually more intrinsically enjoyable. It was true from the time I began to program that I was better at programming than virtually everyone around me. Thus, my intrinsic rewards for programming were higher than virtually anyone else in my peer group, and thus I spent more time programming.

Music is an incredibly clear example of this: while probably anyone can learn a certain level of achievement in music, unless you intrinsically enjoy it, you will plateau quickly. The realization that actually, the intrinsic quality of the person drives how much time they put into their endeavor seriously undercuts Gladwell's argument that the individual qualities of the person are unimportant.

The Good Parts of the Book

The parts of Outliers that I liked the best, and the places where I think it's recommendations have the most weight, is when it veers closest to public policy. It's hard to look at the results in part 3 of the "Marita's Bargain" chapter and not conclude that more school is an unequivocally good thing for children from low-income backgrounds. I find it very hard to argue with the cultural background theory described in the "Ethnic Theory of Plane Crashes."

Though I don't personally like the fact that the "concerned cultivation" parenting style discussed in part 3 of "The Trouble with Geniuses, Part 2" is so much better-suited to our modern world than the older parenting style, this isn't the first book I've read explaining it's advantages.

In Conclusion

In the end, I think the biggest reason the book doesn't sit well with me was its avowed ambition to prove that the personal explanations of success don't work. I found that on the contrary it showed just how important personal drive and ambition are in making use of cultural and demographic advantages. I don't have an issue with the idea that personal explanations of success are insufficient, that we also have to look at cultural backgrounds and demographics as well as the individual to achieve a balanced understanding, but Gladwell aims for more than that.

I think nothing expresses the limits of Gladwell's thesis more than the story of his own grandmother and mother. Gladwell describes the various steps by which his maternal family climbed up the social ladder in Jamaica, with his grandmother taking many steps which eventually propelled his mother and aunt to college overseas. In particular, he notes that the establishment of the all-island scholarships which enabled his mother to attend an elite high school came in 1941, and that were she just a few years older the scholarships wouldn't have been available.

And yet, everything about this story suggests that the key actor in it is not the parliament's establishment of the scholarship, but the grandmother's will to give her daughters every possible measure of advantage. His maternal grandmother is described as a force of nature, and while I have no doubt that the all-island scholarships were a boon for his mother, I also am left certain that the grandmother would have found some other way to get her daughters a high school education had the scholarships arrived too late. To me, this seems to fly in the face of his central argument in the book.

04 March 2009

Convergence is Over

As a Media Lab alum, I have to follow this stuff at least a little bit, so I appreciated Paul Graham's post:

Why TV Lost