Hacker News Reader

And then, assuming you can get people to agree on the accuracy of the observation and on the causation, people can still argue about the best course of action to address a problem. Assuming people don't agree on causation, you might never get enough trials of solutions to learn anything about what a best solution might look like.

Whether or not people agree on something is completely irrelevant to whether or not it is a fact. Earth will orbit the sun, no matter how many people may or may not be comfortable with it.

Finding out what is and what is not a fact can be extremely difficult. However, we've spent the last few centuries developing tools to tackle this problem in the for of the scientific method.

The comment I was replying to pointed out that accurate observations are indisputable facts. To paraphrase the next part, determining causation from observations can be a matter of fact (where you understand the mechanism of causation very well), or of deciding that we're satisfied that the experimentation we've done demonstrates causation (which is not at all the same thing as proving causation). We can also do something like offer a plausible mechanism for causation, which, while often useful, is not particularly close to a fact in the vein of the Earth orbiting the sun.

You're obviously correct that belief doesn't make something a fact, but what to do when a fact hinges on a study? Who decides if the study is good enough? Obviously there is an objective truth, but is the study enough to find it?

> what to do when a fact hinges on a study? Who decides if the study is good enough? Obviously there is an objective truth, but is the study enough to find it?

Good point. This is definitely not trivial or easy - but this exactly the problem that empirical science has been developing tools for.

Technically, a theory cannot be "proven" - at some point, there could always appear another Einstein and overturn everything - but there are still theories which stay consistent with all observations - i.e. "facts" - for an extensively long time. The standard model of physics is such an example: It's not "proven", but for decades any attempt to disprove it has failed and an enormous amount of working technology has been developed by assuming the model accurately reflects reality.

In a similar way, a single study is probably not enough to make a new theory credible - however if the claims of the theory can be consistently reproduced in additional studies, this gives more justification to rely on it. (And more justification to be sceptical of critics)

> Whether or not people agree on something is completely irrelevant to whether or not it is a fact.

No, it's actually the ENTIRE meaning of "fact". Because "fact" is a word, and it has no meaning if we don't agree.

I'm sorry, I don't get it. Could you elaborate?

«Earth orbits the Sun» is not a very meaningful or scientific statement. «Frame of reference tied to the Sun is closer to an inertial one than the one tied to Earth» is better, but try explaining it to random person on Facebook.

What's wrong with "Earth orbits the sun"?

For nitpicky physicists, all objects in a system orbit each other, so the Earth as well as the sun and other planets orbit a common centre (which happens to be inside the sun).

For more obtuse people, they might nitpick that the system should include the entire milky way (so the earth as well as the sun rotate around the centre of the galaxy).

> Example Fact: I observed that in a sample of X smokers who smoked Y cigarettes a day, Z developed lung cancer.

> Example non-Fact: Smoking causes approximately Z rate of lung cancer.

Apart from the specific problems we have ever proving causality. both these statements are still conclusions based to varying degrees on beliefs which may be wrong. The people in the study might have lied. The cancer test might have been flawed. Ultimately the feeling of knowing a fact is an emotion. In science all we can do is be as careful as possible in stating how an experiment or study was performed and what happened. People certainly gain in their confidence of these broader statements as the evidence piles up. But from the perspective of science this reliance on beliefs created by previous studies is a form of bias which can be a bad thing.

I'm not clear on what you intend as a proposed definition, so to help get us on the right track, I'll take a few stabs at distilling down your post to what I think you might be suggesting. Is one of these close?

1. Facts are propositions that are indisputable.

2. Facts are propositions we are certain are true.

3. Facts are directly observable.

4. Facts are empirically verifiable.

Yes. For one, you don't take into account outright lying about results (which is shockingly common) - also by your definition we can never really agree on anything.