1 min read

Specificity and knowing which question to ask

I recently read the back and forth between Scott Alexander and Robin Hanson on whether modern medical spending is effective.  I’m a fan of both authors and I thought the exchange was interesting. That being said, there’s a broader rationality skill I want to pick out that the exchange reminded me of.

I think one of the most important skills when assessing arguments is being able to clearly identify which question or claim is being disputed. It’s okay to start with a less clear general claim and then drill down to cruxes. It’s also okay to disagree about which claim is really important. It’s also true that in reality beliefs aren’t isolated chains of evidence but rather complex interconnected webs and sometimes it’s necessary to challenge/discuss a whole frame/worldview rather than a single isolated claim. Still, in the vast majority of cases being able to clearly pinpoint what claim is being made by someone is possibly the most important skill you can have and looking back at my interactions with various people, I find that a key trait setting apart good conversations from bad ones is whether my conversation partner could understand and be specific about claims or not.

In debating we had a similar concept we called “clash”. There were two competing definitions of clash. The wrong definition was that the clash was the part of the argument space the teams fought over most. The correct definition was that the clash was the cruxiest part of the debate, the key claim or dispute that, if either side decisively proved to the judges, they would likely win. Novices often had to be poked and trained out of being constantly off-clash, that is making irrelevant arguments or wasting time on window dressing instead of disputing central claims in the debate. Watching novice teams debate would likewise often be frustrating as both teams would miss the clash and talk past each other, resulting in “debates” with little engagement where both sides basically gave monologues. I feel that many discussions regular people have are like this.