[mcv]

> How can we disprove the hypothesis “If we can’t disprove it - it isn’t science.”?

That statement isn't science. It's a definition. It's philosophy of science. It's the briefest summary of Karl Popper's definition of the scientific method. According to him, science can never be proven, only disproven.

[ukj]

It is in the nature of definitions to include some things and exclude others.

If the definition of “science” excluded computer science as a science; would you say the definition is correct?

[mcv]

In this context, most of computer science is more a form of applied mathematics.

Of course there are different ways to look at science, like making a distinction between analytical (or empirical) science, and synthetic science; the science that makes stuff, rather than analysing it. Not sure if that's really a good distinction; the latter is really technology, isn't it?

[JohnFen]

There is such a thing as computer science, but the majority of what gets called that is really engineering, not science. People often get those two things confused because they have a fair bit of overlap in the Venn diagram, but they are two different things.

[ukj]

I am talking about (theoretical) computer science, not (practical) software engineering.

This should've been clear in the context of my question:

"Are the formal sciences (logic/mathematics/computer science) not science?"

[mcv]

Math is itself indeed not science. It is the language of science. It follows different rules than empirical sciences. But note that word "empirical" there; Popper was really only talking about empirical science, and according to him, that was the only real science. You could argue that there are non-empirical sciences.

Another problem with Popper is probably that outside of physics and chemistry, there are a lot of less exact sciences where predictions and refutations of a theory are never that clear cut. Like his issues with the theory of evolution.

Ultimately, I guess science is also simply "getting to stuff that works by trial and error".

[ukj]

Ok… I reject your reality and substitute my own.

In this reality I don’t have to do the mental gymnastics where the formal *sciences* aren’t sciences.

In this reality there is at least one unfalsifiable (not even in principle) true claim: the halting problem

This renders falsifiability as a modal criterion. Useful in some scientific contexts - useless in others.

[anonymous_sorry]

> "Are the formal sciences (logic/mathematics/computer science) not science?"

Not by this definition. The distinction between mathematical theorems and scientific theories is a useful one.

[ukj]

>Not by this definition.

Which is precisely the problem with all definitions I am drawing attention to - they are exclusionary in nature.

Sometimes creating distinctions is useful. Sometimes erasing distinctions is even more useful.

How much physics could you do without Noether's theorem?

[anonymous_sorry]

How much engineering could we do without Mathematics? How much commerce?

I don't see it as exclusionary. You won't find many scientists in doubt about the fact that everything they do is built upon Logic and Mathematics, in addition to observation.

But don't we need a word to group fields that try to systematically describe, understand, and make predictions about the physical world? (Rather than seeking to explore and characterise idealised logical constructs?). What would you suggest?