Every time someone posts this, I wonder if they'd be able to solve some basic (albeit symbolic) eigenvalue problems. Could you? Like just some 2x2 problems that you can't just plug into Wolfram Alpha?

It seems a little odd to hold a journalist to a standard software engineers themselves don't generally meet.

I don't think anyone's saying Gladwell is dumb for not knowing about eigenvalues. They're taking his error as a symbol/symptom of his dilletante approach, and the fact that he's making bold claims and 'teaching' people about topics he has no real knowledge or understanding of. Steven Pinker put it this way when coining the phrase:

> the Igon Value Problem: when a writer’s education on a topic consists in interviewing an expert, he is apt to offer generalizations that are banal, obtuse or flat wrong.

He's a writer. Nobody, ever, has confused him for an algebraist. I'm not sure that makes him a dilettante, unless all writers are that.

He writes big, popular books making grand, confident claims about the world, and purporting to back them up with scientific evidence. Experts in the fields he draws from regularly point out significant errors (or misleadingly selective presentation of facts) in his work. The Igon Value thing is a reminder that his work is more a storyteller's patchwork than a researcher's synthesis. (You'll say it's obvious that he's not a researcher; maybe, but some pop-non-fiction writers have more of a researcher's mindset than others. They might not be true experts but they read deeply in search of the truth.)

I mean, it's certainly true that he's not the most rigorous of the pop nonfiction writers! That's not so much the draw, I don't think, but I can't fault you for being annoyed that people behaves as if it was --- perhaps Gladwell himself included.

The point isn't about solving the problem, but at the very least spelling correctly what the problem is.

When I talk to a friend and I'm unsure about something I'm about to say, I try to express my epistemic deficiency (e.g. "if I remember correctly ..."). But when I'm giving a presentation to even 5 people I'll make sure to do a quick google search to confirm if I remember correctly. If I was writing an article that I know will be read by thousands, you bet I would do a lot more.

I guess I just wonder which is worse, misspelling "eigenvalue" in such a way that it's clear that the author is piping their source directly to the page, or spelling it correctly without understanding it, so we have to wonder how much of the mathematical validity of the source the author is vouching for.

Since I don't expect an ordinary writer to have a lot of linear algebra, it doesn't bother me much to know that they're taking their source's word for it on this stuff. But I do worry a lot about the ability for slick-talking people to persuasively name drop concepts they don't really grok.

The context that gets forgotten when people bring this up is that the essay initially appeared in the New Yorker in 2002, where "eigenvalue" was spelled correctly. All the controversy about whether he understands incidental math seems to come from his book publisher having a less careful copy editor than the New Yorker.

Maybe people could blame him for allowing the error to appear in the book, 7 years after it had been corrected in the magazine. Shrug.

Wow, and everyone said that all that math for a CS degree is useless, finally the time has come! /s

So, how come no one caught the error before publication? How come Gladwell did not even look up what an igon value is? Etc.

Everyone makes mistakes, but seemingly the more successful the author the more responsibility they should bear that their overall point is correct and useful. And Gladwell serially misses these marks. At least that's how I interpret these arguments.

I'm with you. The focus on this line also ignores context.

Gladwell is directly quoting Nassim Taleb, and openly says how little he follows the math Taleb discusses earlier in the same chapter. The point is "Look at how smart Taleb is, I don't even know what half these words mean", he was never trying to understand the math or imply he did. In that context mis-transcribing eigenvalues doesn't feel nearly as damning as it's made out to be around these parts?

The point of the "Igon Value Problem" is to have a term to describe popsci writers who create oversimplistic and overconfident narratives of complex domains that they know little about, often making basic errors and leading readers astray.

Whether or not I know about eigenvalues myself is a bit besides the point (although, yes, I did learn that in college).

Well, we're on a message board, so of course you're going to say you know what an eigenvalue is. That's the point of asking if you could handle some simple matrix questions. :P

I still don't understand the point of asking me that. I'm not the one writing books or articles in an authoritative tone on topics I know little about.

If I were to write an article about math with glaring errors, then your question would make more sense. Until then, my linear algebra ability isn't relevant.

Classic - thank you for sharing!