Lung cancer risk was established with control group. We don't have control group for a planet so we use climate models. The article is about remembering that we use models because of lack of alternatives, not because they are as good as control group.

The parent was suggesting causal inference from general statistical evidence to particulars is impossible.

You seem to be suggesting there can be no causal inference where there are no control groups.

Both positions are wrong.

I am saying it is better to have control groups than not. Hopefully that is non-controversial.

That isn’t controversial. The issue is your first point, there is no control group for the Earth as a whole system. So demanding a control group as a barrier to causal reasoning about the relationship between weather and the climate is an impossible bar.

Weather is real. Weather has causes. We can and do reason about those causes all the time.

No I stand by it, talking about cause when the effect is probabilistic is an endless quagmire that people should avoid. Trying to attribute an individual case of cancer to smoking is nigh impossible, which is part of the reason tobacco companies got away with it for so long.

Plenty of companies have gotten away with bad practices by simply demanding proof for each individual case, the earlier we recognize this as a meaningless question the better.

Smoking causes cancer is true. But not everyone who smokes will get cancer and some people who don’t smoke will get cancer. Causation is hard. But we also know that particular cancers are caused by particular acts of smoking. The causal generalization is built on the observation of particulars. It might be hard but I don’t see why we wouldn’t, a priori, be able to say ‘this event was made worse by climate change because it likely affected input x’.

I generally agree with you that this isn’t easy to do, the value isn’t clear, and whether we can say such things with the evidence we have for any particular event isn’t obvious.