1. Factual: These are beliefs based on models, e.g., "the Earth goes around the sun". You can test these models with predictions and if the predictions are wrong, you need to update your model.
2. Value preferences: E.g., "chocolate ice cream is the best dessert". These are non-testable beliefs because they do not have an objectively correct value. You can't say that a preference is "wrong" or "right".
The problem is that all important arguments are arguments about value preferences, but we treat them as factual arguments.
Take something like what to do about global warming. This seems like a factual argument because lots of people don't even agree that the world is getting warmer.
But the real problem is a value preference: People who don't believe in climate change simply don't think the effort required to solve the problem is worth the benefit, and they're using motivated reasoning to deny the warming.
This is obvious when you see them move the goalposts:
1. The climate is not changing.
2. The climate is changing, but not because of humans.
3. Humans are changing the climate, but it's too late to do anything about it.
The real issue is that we disagree on how much sacrifice we should make to help people of the future--people who, on current trends, will be richer than we are, just as we are richer than the world of the 1950s.
But there is no correct answer to "how much should we sacrifice to solve global warming." It's a value preference.
And this is true for every issue. Any issue can be reduced to a disagreement on value preferences.
[Thank you for putting up with my ranting--I just had to get it out.]