Chris Petrilli nails the difference between dynamic languages and the mainstream static ones:

I think that dynamic languages cater especially well to this issue for a few reasons. First, they do not muddle your code with non-core expressions. This means that 80% of my code isn’t spent making the compiler happy, or doing its job for it. The code that I write is focused purely on the problem domain that I’m trying to solve. This means when I go back, or anyone else does, there’s less time spent trying to understand why I marked something as final and more about what the logic does.

Couldn't have said it better myself. Who are you looking to satisfy: the requirements, or the compiler? The more of the latter that you have to spend time on, the less you end up with the former.