Dynamic inspection of an object's API? I think it's a very common thing for programmers to explore a type's API on-the-go (as we can't memorize every function name). With static typing, that information is known before run time and can be immediately displayed by the IDE in the form of auto-complete. In dynamically typed languages, the most reliable way to achieve the same is by running and inspecting it, which is basically how people work in Smalltalk from what I've seen.
Things are a little confusing when we can both use the Smalltalk IDE with or without running our stuff.
Code exploration tools are available before "running and inspecting" our stuff. We read the protocols and code without "running and inspecting" our stuff.
It is true that static types provide more-precise more-automated identification of implementations and call sites.
