I'll looking at the parser example, page 242 in Chapter 10 of Real World Haskell, and they are defining a type of monadic parser with the help of an operator they call ==> You can find chapter 10 online. This ebook doesn't have page numbers, but you can find the example I'm looking at in the second called "A more interesting parser", about 40% of the way down: http://book.realworldhaskell.org/read/code-case-study-parsing-a-binary-data-... The authors have defined their parser by chaining together functions with ==>. The first function is "getState". What confuses me is: they use getState to "get the state out of the Parser," but a Parser is by definition a function that takes the parse state as its lone argument. I don't understand why they can't drop getState entirely. Maybe it's a simply a way to avoid wrapping the entire function in Parser (...)? Some of this stuff looks "inefficient" to me, but I realize that in a lazy language with an optimizing compiler you can often write long chains of functions (many of which discard their results) and not impede efficiency. Thanks, Mike