Taking a simplified model

This may be trivial but I didn’t see it written down explicitly anywhere, so I’m writing it. There’s a specific trick I’ve noticed that some people do that makes for better conversations, and I don’t know if there’s a name for it, but I think there should be, so I nominate taking a simplified model.

Say that you want to figure out some big societal issue, like how taxes should work, or how much people should be paid, or how much of a company should belong to the founder, and how much to the workers.

The usual way to go about this is to launch into specifics or into generalities. You either talk about the tax situation in, I don’t know, post-war Netherlands and how the price of oil affected the shipping industry which then led to something which led to something else - I’m just fabricating stuff, but you know, a line of reasoning where you extract some lesson from highly specific events.

Or you talk about very general principles - like, we should all be treated equally, or we should reward those with ingenuity and drive, stuff like that.

But the cool trick is taking a simplified model: e.g. imagining a society of only ten people. And then running through the combinations, maybe extracting lessons. For example:

• Imagine that in this society, John accidentally discovers oil (or an imaginary resource, doesn’t matter). John works hard and makes life better for the rest of the society. How should he be rewarded (if at all)?
• Imagine that John doesn’t work very hard, but is just the first person out of many who might have plausibly discovered this new resource. Does this change anything?
• Imagine that John works on this project single-handedly but wants to leave his heritage not to the village but to his two sons. How do we feel about that?

And so on. This works for a broad set of questions, and I’ve found that taking a simplified model always brings clarity. The trick is that it doesn’t work in highly soldier-like debates, where such hypotheticals might be perceived as gotchas. But when trying to figure things out, reducing the number of variables works really good. You answer a simpler question, and only then make the question more like real life.