What do Prototypes Prototype?

I was reading this paper on Prototyping, titled "What do Prototypes Prototype" and found the whole scenario to be really interesting.

Prototypes are such an interesting space. Ofcourse you can't build the thing at once, or else it'd be a fully functioning product at that point. So it must be something that's in-progress, but how does one show that. and when to "fake" something vs when to show something as real is such a pertinent question.

To me one of the best advice I've received is, everything should be fake except for the thing you're testing the prototype for. That should be real. So then it really becomes the question of "What answers we need from a demonstration", and then that informs what is prototyped.

Wish I knew this before.

BTW, this also hits at the problem of making first, or doing research first. If you look at it in one way, prototypes are ideas put out of the brain in tangible form. Making a prototype should be non-negotiable when showing something to people. Understand a problem space, yes. But don't do research without a prototype, would be my advice.