Yes, you can. I'm not a Haskell guru, and I haven't played with it in some time. I can read that line of code just fine. I don't actually know what it's doing, because I don't know the functions involved, etc., but the form of the line is easy enough to grok. Like any other language, it looks mysterious if you haven't read and written enough code in it, or in a language with a similar style.
In a very real sense, that's true. But...I don't know what it's doing but I know how it's doing it, so to speak. With languages in which I'm quite fluent (Haskell not so much), I can sometimes spot problems in the code of someone coming to me for help without knowing anything about the libraries, et al, because even though I "couldn't read it" in your sense I could read it in the other sense.
Can you read that code? Yes? But you don't know what it is doing, because you don't know what the functions foo and bar do. Same deal with the haskell code in question. Just because you don't know what Compose and getCompose do, doesn't mean you can't read the code.
I don't feel it's so clear cut. Haskell fuzzies the difference between syntactic and semantic elements. What in your piece of code is the parentheses might very well be defined as functions in Haskell. For instance, the ($) and (.) functions, but also I would call StateT and perhaps Compose and WrapMonad (while not knowing what they do) such things too.
So, I still feel that understanding a piece of Haskell means that you will need to understand the individual functions at least superficially.
You feel wrong. There is no such fuzzy. StateT, Compose and WrapMonad are functions. They are just like foo and bar in my example. Just as in a C-style language you need to know the parens mean function application, in haskell you need to know spaces and ($) are function application, and (.) is function composition. The code is perfectly readable even if you don't know what the functions being used do.
