Joel on Software: Anecdotes
joelonsoftware.com
joelonsoftware.com
She mentions only one of the many larger studies that he cites, the Terman study. Was the Terman study wide ranging? Yes. Does it fit Gladwell's hypothesis that the opportunities one is offered matter? Yes. Does she show it to be incorrect? No. Does she attack any part of it? No. Does she mention any of the other studies that Gladwell mentions? That major hockey players are nearly all born in the first three months of the year? Ericsson's studies on expertise? Flynn's analysis on historic chinatowns? No!
If I present a theory, and back it up with a bunch of other data points, one should not be able to pick out a single data point, claim that I am basing my loose hypothesis on anecdotal evidence, and say I am irresponsible. This is what is being done.
You've pretty much described 95% of the fun of reading the New York Times' book review. It's a public forum for less-famous writers to snipe at more-famous ones -- it's practically the first blog.
I've read the book. It's entertaining, but I don't think it's as well-developed as Blink or The Tipping Point. I think Gladwell is a good writer who serves a previously unfilled role as a popular popularizer of academic work. However, every time I read his work, I wish that experts of the fields he reports on would also write popular expositions that could be consumed by an audience like that of the New Yorker.
I know that the evidence backing some of what I was reading was sketchy at best. Asian children score higher on math tests because their ancestors labored in rice fields? Please. It doesn't detract from the entertainment value of the book, but I would think twice before believing what I read in the book was indicative of reality.
As for your theory argument, I would have to disagree in many different cases. Suppose that you only chose to include data points that support your argument. Then people have every right to pick an alternate data point, claim that your study is faulty and that you're being irresponsible. This is exactly what many of his detractors are pointing out, and it is a point that is often missed.
The primary argument against the book and the Gladwell style of investigation is that he tends not to include or even discuss studies or points or examples that do not support his argument. This is understandable, since he is attempt to speak to a popular audience and does not have the space to discuss every relevant study if he wants to run through all of his cases, but by doing this, he makes many smart people feel like he is trying to pull the wool over their eyes.
I wish that were true too. I am trying in my own blog to present my work in both a form accessible to the literate public and helpful for researchers (though I have neither published most of my work yet, nor attracted an audience the size of the New Yorker so you will just have to be patient!)
Arguably, Flynn did his best with "What is intelligence?" but people probably ran for the hills at the first formula. Such is life. But there are at least a few scientists who write for the public audience, Dawkins, Dyson, Hawking and Pinker perhaps most relevantly today, that do attract a wide readership. I honestly think Gladwell does a better job presenting things straight than Hawking.
The rice field argument, while apparently silly, doesn't strike me as particularly wrong. I'm merely unconvinced.
Does Gladwell in this book, more than his others, draw theories from selected scattered 'outliers,' more often than reasonable? I do not know how far he goes.
Jeff Atwood took a lot of flack in his recent NP post simply because he dared describe the problem in non-mathematical terms. He may also have been wrong, but I am inclined to think his explanation was (correctly) too simplified to really be wrong... but still, right now I'm talking about the people who flamed him for failing to include a complete computer science textbook in his blog post.
Yes, I am so sorry that in my blog post ruminating on how unit testing seems to affect my Python coding that I did not include a complete, correct Grand Unified Theory of the universe complete with how to apply it at the macroscopic level and use it to predict who will win the Presidential election in 2016 and create a superhuman-intelligent AI. At some point you to understand that any given work has boundaries.
I've become sensitive to this because I actually have fallen for the trap of trying to be too inclusive, and it's worthless. Stop complaining about it. Or, at the very least, when complaining that work X should have included Y, you should say what should have been cut instead and why it would be an actual improvement in the end product, preferably beyond "I would like it better".
I think that's why many of these books are so frustrating to read -- they seem to rely almost completely on anecdotes, without any sort of underlying data to look at long-term/overarching trends/themes. Thus, instead of saying, "here are various data, including these representative anecdotes about some of them, and so we can see that my theory is true," these writers usually say "this is my theory, and look how well these carefully chosen anecdotes match it!"
If you give us the anecdote, but don't extrapolate to wild grand theories, you're doing it right... simultaneously telling an interesting story and giving people more data points to build up their own understanding of how the world works.
If you give us an anecdote, then tag on some theory so that the anecdote isn't mere storytelling, trying to make the anecdote more than it really is, it's deception (at best), or, at the very least, talking about something you know nothing about.
And all these anecdotes and know-nothing books, blog posts, and articles are crowding out real, genuine stories and useful knowledge.
Terman et. al's studies of gifted children (http://books.google.com/books?id=MXwVAAAAIAAJ&printsec=f...)
Sherar et. al's study of physical maturity in hockey player development (http://www.informaworld.com/smpp/content~content=a778078133~...)
Flynn's book "On Intelligence" and review of the growing IQ of the worldwide population and ethnographic studies of Chinese. (http://books.google.com/books?hl=en&lr=&id=qvBipuypY...)
I don't think one should claim an insufficient justification of his hypotheses without actually examining the justification thoroughly.
Oftentimes, for Gladwell in particular, people claim that he's being unrigorous, imprecise, or incomplete. He's not being incomplete. He's being concise. If he weren't concise, almost nobody would pay any attention. Almost nobody would learn anything from him. And almost nobody would, wanting to learn more, have the opportunity to follow the reference list that he carefully puts out. Which I guess few people do, anyway.
Since it's pretty so rare to find someone who both (1) spends the time to find and understands that work, and (2) has the literary ability to engagingly present it, by some definitions, that can be considered genius.
Still, something about both Friedman and Gladwell really bothers me. I think it is simply that I believe their theories are mostly wrong or useless. It's not the style or methodology that is the problem, but the results.
Every single one of us is a collection of anecdotes, personal experiences and gathered data. Every single viewpoint ever held by every single person is a theory based on such a collection.
You aren't honestly suggesting that those opinions that are based more on the first two than the latter are completely invalid, thus unworthy of being shared?
Sure, books like the ones you describe - and in fact I'd posit at least 80% of your own blog posts, if not more - are more heavily invested in anecdote than they are proper data. They, and the conclusions they present are nonetheless valid pieces of information and help build a balanced view that is continually evolving.
The world is simply too complex a place to gather hard data on every possible subject matter before forming an opinion or articulating a theory. Some of the most outlandish theories in history - evolution, string theory, climate change all come to mind - start from anecdote and observation prior to being supported by hard facts and abundant data.
There are different stages of the journey, I suppose.
Obviously, it's best if an essay is both truthful and entertaining, but if an author decides to trade scientific precision to achieve a more vibrant story, that's fine with me. The world is far too big to stick to the facts.
Gladwell isn't a scientist, as he has said many times. He is inspired by scientists (among other people), but what he actually writes are book-length essays. You're welcome to argue against an essay -- that's what the form is designed to provoke -- and you can do so in several ways: You can write an essay of your own, or you can do a scientific study, or you can even perform a stunt:
After we came out of the church, we stood talking for some time together of Bishop Berkeley's ingenious sophistry to prove the nonexistence of matter, and that every thing in the universe is merely ideal. I observed, that though we are satisfied his doctrine is not true, it is impossible to refute it. I never shall forget the alacrity with which Johnson answered, striking his foot with mighty force against a large stone, till he rebounded from it -- "I refute it thus."
But the modern fashion is to respond to an essay that you don't like by demanding that it dress itself up like a Nature manuscript and then criticizing the flaws in the costume. Which is silly. The more elegant thing to do is to argue with the essay. Write an essay refuting Gladwell. Quote some scientists of your own in the process if it helps. Obviously, debating a guy with the writing chops of Gladwell is a tricky task, but it's better to try (or to remain politely silent and wait for others to try) than to flail around ineffectually.
Or, if you prefer scientific debate to essay-writing, I'm told that Gladwell seems to be trying to provide more footnotes to the science that inspires him. Follow one of those footnotes to the scene of battle, grab a handful of related journal articles, and join the fray. Don't be surprised to find every other scientist on earth alongside you, though. Scientists have a tendency to swarm fields which have been popularized.
Essays make you generalize; they're short. But generalizations need interpretation.
And if you really want to know something, you read a technical paper or do it yourself.
I could go on about this in comment but won't; Comments want to be witty.
Though unlike Chris Anderson, at least the anecdotes Gladwell bases his stories on are true, rather than mistaken third-party analysis.
We need this variety of uninformed opinion because it is from one of those nuggets that new things will be invented.
Science is just philosophy combined with statistical data.
Irony, defined. :-) I kid.
I think Joel's problem here is that he missed the point where any of these guys claimed to be doing anything more than making a bunch of observations. Galdwell certainly isn't science, but I don't think he proclaims to be either.
I haven't read his newest book, but certainly his thesis is at odds with the idea (the one we truly want to love) that the only thing separating the haves from have nots is desire and work ethic. We're all over saying this is common sense, but then we find it outrageous that this could have an affect on any one particular outcome.
Gladwell's ultimate theory I suppose is that his observations should make us question how our society functions overall (again, haven't read the book yet). I guess that forces us to look at things in a way we often don't enjoy.
Well I'm going to compound it by adding my observations :)
I've got to say, in my own experience, I like anecdotes that are war-stories best. A critical self-analysis often is full of insights for others... The most gripping, amusing and insightful talks I've seen have often been "here's how a large project detonated on us".
Unfortunately, in my experience (I end up at a lot of SOA, Cloud, Web/2.0 type events) these are quite rare for a variety of reasons . You tend to get more of the hypothesis/external critique - which by and large I find a lot less valuable.
Anyway, that's my anecdote on anecdotes :)
The problem is that the majority of people will think it is science. It is written in a book, so it has to be true.
Would someone care to define this proper science?
Usually to call something "not science" is to say it is not based on 1) testable, falsifiable hypothesis, 2) accurate data, or 3) an inability to replicate/verify
Books like those you see in the airport, the type that Gladwell, Godin, and the like write, are designed to be science-entertainment. It's just science (or marketing, in Seth Godin's case) as entertainment.
If these sorts of books actually get people interested in and reading about science in the first place, it's awesome! There's no way the general public will read dry academic papers, no matter how scientifically valid they are. If people build up an interest from somewhat exaggerated but entertaining books that introduce certain scientific principles, we are improved as a society.
But the plural of 'Anecdote' is not 'Data'.
That's the problem: interesting data points become anecdotes while uninteresting data points don't, which is why you can't reconstruct the data from the anecdotes.
A plurality of data sampled without bias?
That's the problem: interesting anecdotes become data while uninteresting anecdotes don't, which is why you can't reconstruct the anecdotes from the data.
1. We have just spent an entire presidential campaign during which we were reminded at every turn (and rightfully so) that the gap between rich and poor is growing, and this seems to dovetail nicely with a "theory of social predestination." Having anecdotes and data to try to explain the mechanism for how this works seems a worthwhile undertaking.
2. Why does she want Gladwell to research high speed Internet connections?
Joel written many an essay with sole basis in anecdote.
This is what has always bothered me Gladwell's writings, and when I see someone list 'Blink' as one of their favorite books it pretty much tells me they are someone of little substance. Gladwell is simply a formula writer (and his speeches follow a similar format) and I find it amazing he has extended his popularity this far.
So suddenly the software world's number one anecdotal blowhard decides to slam another blowhard. Okay. Whatever.
So bottom line, he really flatters successful people on the surface, while making the successful folk equal to the rest of us.
I don't think you are "mad" at Gladwell. I just think the motivation is strong and possibly an unconscious aversion to accepting that your success may be in huge part due to luck.
But then. Take with a huge grain of salt some asshole spouting off on a forum.
Believe it or not, you are one of my heroes, and I never seriously expected YOU of all people to be reading this.
Cheers to you and don't stop doing what you do. We are a better field because of it.
... and thus... Squeezed Books :-)
So people like 37signals had a good pinch of luck thrown in their pot, but Joel, I believe, actually got to the top rung of the internet food chain by persistence and hard work.
1. When someone "successful" criticizes Gladwell's thesis, it's suspicious because maybe they just don't want to hear that they weren't really smarter than everyone else.
2. When someone "unsuccessful" criticizes Gladwell's thesis, it's suspicious because maybe they just don't want to hear that their lack of success simply shows that they didn't work hard enough.
#1 is your argument; presumably you find it plausible. It seems to me that #2 is about equal in plausibility to #1; and #1+#2 would say that anyone should be viewed with suspicion, as probably motivated by something other than honest intellectual inquiry, if they criticize Gladwell's book. Which seems ... unhelpful.
It also seems curious that you describe someone as (a) "the number one anecdotal blowhard" and (b) "one of my heroes". Why should the rest of us take any notice of someone who knowingly takes an anecdotal blowhard as a hero?
Joel Spolsky: Hard work + good luck = success! You could be Gladwell's next anecdote.
I think you can see the effects of luck and diligence in just about everyone's life. That's not at issue. Gladwell is suggesting that to be a Bill Gates requires a good amount of diligence (10,000 hours of practice) and also utterly ridiculous good luck. That is, the skill component stops scaling really quickly, but there's no limit to how lucky you can be.
The critics are proclaiming this to be both unsubstantiated and a truism. But I think Gladwell is onto something, because certainly our culture treats successful people differently. Think about all the successful businesspeople invited to Davos. Maybe future generations will look at that and wonder what we were thinking. Like we gathered together a lot of people who won the lottery, in the hopes that they'd win the lottery again?
Aside: watch for when people say "whatever." It means they care very much about something.