Sunday, December 6, 2009

Gladwell's "What The Dog Saw"

I have been reading Malcolm Gladwell's book of essays for a while now.  Reading these articles reminds me of why I like Gladwell and why he drives me crazy.  He asks really good questions, investigates them in interesting ways, but often provides weak answers.

To illustrate, I will first summarize his first three books:
  • Tipping Point:  It basically ponders why things spread.  The book argues that it has to do with the nature of the thing, the attributes of the messenger, and the context.  So, in some ways, everything matters, which is pretty unsatisfying.  But in the course of the book, Gladwell reviews heaps of interesting social science.
  • Blink: Your first intuition is correct except when it is not.  Least satisfying of the three books because its conclusion is so banal.
  • Outliers:  What causes success?  Context and hard work, not just talent.  The fun stuff here is the context stuff.  That the best hockey players in Canada tend to be those that were born in the first three months of the year, as they would have a physical advantage over the other kids and that would lead to more reps, more coaching, better competition, etc.  In other words, path dependence is huge here.
 I am not going to go through each and every essay, but highlight some of the more fun/provocative:
  • "The Pitchman"  Documents the tales of the folks behind Ronco and other kitchen gadgets.  Just a fun bit of history behind the infomercials.
  • "The Ketchup Conundrum"  Explains why there is just one dominant Ketchup company and many flavors of mustard.  And how yellow mustard was supplanted by fancier forms.  A nice lesson in the science of taste and how Heinz more or less accidentally found the perfect formula.
  • "John Rock's Error"  History of the birth control pill, with the assertion that it did not have to be a monthly rhythm.  Indeed, heaps of interesting stuff on periodic periods.
  • "Open Secrets"  A tale of Enron, but addresses an important problem for those who seek intelligence--info about their adversaries--too much info can be as bad as too little.  Not a new idea but well illustrated.
  • "Million-Dollar Murray."  Uses the stories of homeless people to illustrate a very interesting problem--most of our work in public policy assumes that the population or problem is distributed by a normal distribution (a bell curve), but power-law distributions tend to mean that most of the stuff is located at one tail or another.  This has powerful implications for social policy, as helping the average person or even most of the people may not make much of a difference since most of the impact (the costs incurred by the society due to homeless individuals is largely due to a small number of the homeless) is concentrated in the extreme ends.  So, policy needs to be re-aimed bigtime. 
  • "The Picture Problem"  Addresses the problem of interpretation, especially of complex visual stuff, like mammograms.  This is apparently more art than science, and I was reading this essay when the mammogram protocols were in the news again.  The challenge here is that false positives are incredibly expensive in terms of medical costs but also emotional ones. 
  • "Something Borrowed"  A fascinating story about how he was plagiarized and that he didn't mind that much since it was for another medium--a play.  The funny thing is that he person he profiled in the original piece was far more upset about the plagiarism.
  • "The Art of Failure"  An interesting discussion of choking vs panicking.  But it does not really tell us why either happens.  Choking refers when people lose their instincts and end up relying on the formal lessons they learned. Panicking is "the reversion to instincts."  Fascinating but frustrating--perfect Gladwell.
  • "Blowup"  Summarizes Perrow's Normal Accident stuff.  Some hints of the Traffic stuff by Vanderbilt is here.
  •  "Most Likely to Succeed"  The problem of hiring people when we really do not know who becomes a good x.  Like Quarterbacks and teachers.  I was going to suggest that we have the same problem in academia, but probably not as severe. 
  • "Dangerous Minds"  That the profiling of serial killers we have heard so much about is, well, perhaps a bunch of hooey.  The original studies were very much products of selection bias, and then Gladwell compares the profilers to huckters--to Mentalists (which was heaps of fun, given my love for the show).  Gladwell cites a variety of tricks from Ian Rowland's The Full Facts Book of Cold Reading: The Rainbow Ruse (where a person  is both x and not x), the Jacques Statement, the Barnum Statement ("the assertion so general that anyone would agree"), and so one.  Fun, fun stuff, and really challenged what I had previously believed.
 I have two more essays to go.  Will update when those are done.


Jacob T. Levy said...

The Ketchup Conundrum is such a great essay that it earns Gladwell forgiveness for at least a few bad books.

Steve Greene said...

The Ketchup article changed my life. I've been all about Umami ever since.