What's the point of map in Haskell, when there is fmap?

Everywhere I've tried using map, fmap has worked as well. Why did the creators of Haskell feel the need for a map function? Couldn't it just be what is currently known as fmap and fmap could be removed from the language?