How about: -- | Splits a list into components delimited by separators, where the -- predicate returns True for a separator element. The resulting -- components do not contain the separators. Two adjacent separators -- result in an empty component in the output. eg. -- -- @ -- > split (=='a') "aabbaca" -- ["","","bb","c",""] -- @ split :: (a -> Bool) -> [a] -> [[a]] split p s = case rest of [] -> [chunk] _:rest -> chunk : split p rest where (chunk, rest) = break p s -- | Like 'split', except that sequences of adjacent separators are -- treated as a single separator. eg. -- -- @ -- > tokens (=='a') "aabbaca" -- ["bb","c"] -- @ tokens :: (a -> Bool) -> [a] -> [[a]] tokens p = filter (not.null) . split p 'split' is the same as Python's split. There is already a precedent for using short names in the Data.List module so I think 'split' is the right choice, and 'tokens' is fairly descriptive. Then we have words = tokens isSpace which is quite nice. Unfortunately lines /= split (=='\n') because lines strips off blank lines at the end. So perhaps we do need the third version. Cheers, Simon