In the latest Vanity Fair is a brilliant piece of journalism, Goodbye to All That: An Oral History of the Bush White House by Cullen Murphy and Todd Purdum. In a fun, easy-to-read format, it tells some basic truths I had never read before. Here are two examples:

Matthew Dowd, Bush’s pollster and chief strategist for the 2004 presidential campaign:

When Abu Ghraib happened, I was like, We’ve got to fire Rumsfeld. Like if we’re the “accountability president,” we haven’t really done this.

We don’t veto any bills. We don’t fire anybody. I was like, Well, this is a disaster, and we’re going to hold some National Guard colonel responsible? This guy’s got to get fired.

For an M.B.A. president, he got the M.B.A. 101 stuff down, which is, you know, you don’t have to do everything. Let other people do it. But M.B.A. 201 is: Hold people accountable.

David Kuo, deputy director of the White House Office of Faith-Based and Community Initiatives:

There’s this idea that the Bush White House was dominated by religious conservatives and catered to the needs of religious conservatives. But what people miss is that religious conservatives and the Republican Party have always had a very uneasy relationship. The reality in the White House is—if you look at the most senior staff—you’re seeing people who aren’t personally religious and have no particular affection for people who are religious-right leaders. Now, at the end of the day, that’s easy to understand, because most of the people who are religious-right leaders are not easy to like. It’s that old Gandhi thing, right? I might actually be a Christian myself, except for the action of Christians.

And so in the political-affairs shop in particular, you saw a lot of people who just rolled their eyes at everyone from Rich Cizik, who is one of the heads of the National Association of Evangelicals, to James Dobson, to basically every religious-right leader that was out there, because they just found them annoying and insufferable. These guys were pains in the butt who had to be accommodated.

This is related to the Shangri-La Diet. In these two excerpts, the speakers were (a) close to the events they describe but (b) not so close they are in any danger from the people they tell the truth about.

In science the same thing happens. Saul Sternberg and I could tell the truth about Ranjit Chandra’s research not only because (a) we were fairly close to that research (which involved psychology, even though Chandra was a nutritionist) but also because (b) not being nutrition professors, Chandra couldn’t harm us. Those closer to Chandra, professional nutritionists, had plenty of doubts as far as I could tell but were afraid to say them.

Hal Pashler and I could criticize a widely-accepted practice among cognitive modelers because (a) we were in the same general field, cognitive psychology, but (b) far enough away so that the people we criticized would never review our grants or our papers. (Except the critique itself, which they hated. After the first round of reviews, Hal and I requested new reviewers, saying it was inevitable that the people we criticized wouldn’t like what we said.)

Likewise, in the case of voodoo correlations, Hal is (a) close enough to social neuroscience to understand the details of the research but (b) far enough away to criticize it without fear.

In the case of the Shangri-La Diet, I was (a) close enough to the field of nutrition that I could understand the research but (b) far enough away so that I could say what I thought without fear of reprisal. Nassim Taleb is in the same relation to the field he criticizes. Just as Saul Sternberg and I knew a lot about the outcome measure (psychological tests) but were not nutritionists, Weston Price, a dentist, knew a lot about his outcome measure (dental health) but was not a nutritionist.

It’s curious how rarely this need for insiders/outsiders (inside in terms of knowledge, outside in terms of career) is pointed out. It’s a big part of how science progresses, in small ways and large. Mendel and Darwin were well-educated amateurs, for example. Thorstein Veblen wrote about it but I haven’t read it anywhere else.