Crosscut

Traditionally, source code is expressed as plain text. This has the advantage that we can use existing editors and version control tools to manage it. But this plain text form requires a lot more processing, before it can be executed.

And not only that, it might also require a lot of processing before it can be understood easily by the developer. A prime example of that is type inference. Figuring out which type a given value has, might require a full understanding of the surrounding code. But once type inference has happened, your tools can just show this to you.

Modern development environments tend to do this by at least partially compiling the plain text code, until they can provide this information and show it to you. When you make a change to that plain text code, they compile it again (maybe using a cache for parts that didn't change), and show you updated information. It all starts with this plain text code. That's why it's called source code.