« It's a New Phil, Week 56 | Main | A Bipartisan Future »

Foxes vs. Hedgehogs

Phil Tetlock, speaking in a seminar for the Long Now Foundation, gives some thoughts as to why politics tends to make people dumb. He asserts that there are some trans-ideological benchmarks that we can look for in how people think about political issues that enable better predictions.

Tetlock is a psychologist who has been doing research in prediction for some time. He became interested in how political thinking relates to prediction around the end of the Cold War. During the Reagan administration, many liberals believed that Reagan was risking nuclear war (and massively wasting resources on a military build-up) when the Soviet Union was a highly stable entity, much more likely to engage with us in a nuclear holocaust than to back down or collapse. And, in fact, many conservatives believed that Reagan was right to confront the Soviets not because they thought we were going to win, but simply because they believed the Soviets were here to stay and we needed to keep pushing back against them in order to maintain our own credibility. The dominant view was that nothing good could happen.

When the largely unexpected outcome of Soviet collapse occurred, Tetlock observed that no one seemed to acknowledge that they were wrong in their thinking. He refers to this as an outcome-irrelevant learning situation. It then occurred to him that someone should be keeping score on these things. His research is all bout identifying a way to keep score. A major learning from the research is summed up by a quote from the Greek poet Archilochus:

The Fox knows many things, but the Hedgehog knows one big thing.

As Tetlock explains is, one can be a Marxist hedgehog or a Libertarian hedgehog. The content of the ideology is not important. The important thing is that you approach current events with a few limited principles that you want to use to explain what you see. Foxes, on the other hand, take a scattered approach, bringing a lot of different principles into play. Foxes are skeptical of grand ideas, and skeptical of their own ability to predict things. But they are, generally speaking, much better predictors than hedgehogs, especially in the short term.

What do hedgehogs do well? They assign higher probabilities to big changes that do, in fact, occur. For example, those who predicted the downfall of the Soviet Union were hedgehogs. But they are right at a high price. You get a lot of false positives. Hedgehogs were also predicting the downfall of, for example, Canada. Others were predicting Dow 36000, a neo-Stalinist coup in Moscow that would bring back the USSR, and lots of other big stuff that never happened.

He explains that reality is not a horse race between the two viewpoints. Foxes and hedgehogs need each other. Foxes are intellectual scavengers who pick bits and pieces of big hedgehog ideas. They tend to be contrarian. They annoy people across the political spectrum.

Tetlock makes some interesting comments about the War in Iraq as it relates to the subject, and I would have liked to hear more on that. During the ensuing Q&A period, it becomes clear that the audience pretty much thinks that hedgehog=conservative, which is at odds with the basic premise. But overall, it's fascinating stuff. Give it a listen.



Which side bases nearly all of its analysis on the single analogy with Vietnam?

Of course, there are plenty of conservative hedgehogs, and plenty of liberal ones, too. Tetlock's idea that politics makes one dumb seems to derive from the tendency of political ideologies to push one's thinking in hedgehoggy directions.

The quote about hedgehogs and foxes is not from an ancient Greek historian, but from an ancient Greek poet named Archilochus.

Good catch, Peter. I will correct that.

"For 'tis thy friends that make thee choke with rage."

I link over and riff on the possibility of the "fodgehog" over on my blog. But is that ME being a hedgehog about applying synthesis in cases where it might not apply?

BTW, Boxing Alibiades' post can be found over at Happycrow's Eyeball Factory. I usually let readers do their own link-whoring, but what they hey: it's Valentine's Day.

I believe the Fox, in Tetlock's view, is something of a fodgehog in his own right. Foxes do out-predict hedgehogs in the short term, but they also out-predict them in the long term. I think they're already doing the synthesis you're talking about, at least to some extent. Still, hedgehogs add value by (once in a while) predicting a big, sweeping change that really does come about.

Hey, no linkwhoring here... you'll notice I didn't include one in my reply! :)

What I miss from the "fox = fodgehog" is that foxes are uncomfortable staking out clear positions. And that's part and parcel of the game. It's real common for me to give a $2 answer to a dime question on these sorts of things... foxes clearly out-predict. But they out-predict to a very shallow depth. The hedgehog goes further, but becomes dogmatic, as opposed to changing his opinion on a dime in the presence of new data... and therefore I'm not so sure that either functions as the hybrid, able to *flexibly* stake out clear theses.

Post a comment