In math we have an proof technique called reductio ad absurdum that deals with this situation. They seem to be doing it wrong though, because they don't reject the faulty hypothesis after the conclusion is logically inconsistent with reality.

I can apply the same argument to science: all of the laws of physics are symmetric with respect to time and space. But all of the data I have direct access to comes to me from a privileged frame of reference that I call "here and now". Therefore, all of this data is in direct conflict with physics, and so physics must be wrong.

Everything starts with an assumption. However, in science, if the assumption's conclusion doesn't match the observed evidence, we reject our assumption as true, and we start with a new assumption.

As so, science doesn't have a "foundational" assumption (because it's willing to change its assumptions). Rather, science subjects assumptions to evidence - not the other way around...

> science doesn't have a "foundational" assumption

Yes, it does: it assumes that evidence is a reliable guide to truth.

Nonsense.

"evidence" and "truth" are very semantically related. By semantics, "evidence is a reliable guide to truth" is not nearly an assumption... as it is a logical definition.

It's a bit like saying: "it assumes reddish color indicates red".

EDIT: "Reliable" is the problem word here. Remove it from your statement, and it certainly is no longer an assumption. That said, I don't see where science inherently makes the assumption that evidence is reliable...

> it is nearly a logical definition.

Your having to hedge with "nearly" is an indication of the weakness of your position.

First, there is no such thing as a "logical definition". There are only definitions. Perhaps you meant "logical tautology" but in fact the idea that evidence is a reliable guide to truth is neither a definition nor a tautology, it is an assumption. And it is an assumption that is open to reasonable doubt. To wit:

> it assumes reddish color indicates red

It is far from clear that this is true. What does "reddish color" even mean? How can you define it without begging the question? How do you know that your subjective perception of "reddish color" is the same as mine?

A better example: solid objects are in fact not solid, they are mostly empty space. So the assumption that "solid appearance indicates solid" is actually false.

It's important to understand that definitions are not assumptions. Definitions and assumptions are two different things. Now, we could also define them to be the same thing if we wanted, but typically definitions aren't about truth. They're about meaning.

Now, if you'd rather use "logical tautology" that's fine by mean. It's close to what I meant. "Evidence" and "Truth" are very semantically related... so much that your statement earlier is almost a definition, not an assumption. By definition, Evidence is something that indicates truth. Again I think the word reliable does make the statement an assumption. But again, I don't see where science ever made the assumption that evidence is reliable...

As for your second example, it is not much better. I will again attempt to show you the problem with your argument.

>solid objects are in fact not solid, they are mostly empty space. So the assumption that "solid appearance indicates solid" is actually false

For your argument to hold, your definition of solid in your second statement must be incongruent. I'll try to show how.

> solid objects are in fact not solid, they are mostly empty space

Here you've tacked on some definition to the word solid. That added definition being: "they are mostly empty space". Now let's look at your next statement

>So the assumption that "solid appearance indicates solid" is actually false

This only logically true if you use two different definitions of the word solid. Otherwise your statement is nonsense, when substituting your implied definition of solid:

something that appears as not empty, but is in actually mostly empty space appearance indicates something that appears as not empty, but is in actually mostly empty space is actually false