Why does Char have an instance for Bounded?

Character encodings are tricky. Behind the scenes, all characters are represented by numbers. The Unicode standard provides a set of "code points" which are simply numbers which map to a particular sequence of real characters. Unicode defines code points between 0 and 1114111 and so that's what you see when you try maxBound.

Char encodes Unicode code points as individual integers, which is somewhat inefficient. If you want an efficient encoding, use Text.

You're seeing \1114111 displayed because that's the code point that maxBound :: Char represents and there is no more efficient, meaningful way to display it. In particular, it's in the "Supplementary Private Use Area-B" of the Unicode standard which means that it's reserved for use outside of the scope of Unicode and thus has no standard meaning.

All characters, like all things in a computer, are ultimately just numbers. Char represents unicode characters, which are represented via numbers. You can convert between Char and Int values with ord and chr. E.g. the unicode value for a is 97, so ord 'a' is 97 and chr 97 is 'a'.

Char '\1114111' is the Char that represents the number 1114111, or 0x10FFFF, which is defined as a noncharacter. This is the largest value that is defined in Unicode, and is the largest that Haskell supports: '\1114112' will cause a compile error.