Remove duplicate definition of identifiers from WIT.md (#276)

WIT.md had an outdated definition of identifiers, which only allowed ASCII and included the old method of putting identifiers in quotes to escape keywords.

This removes that definition in favour of the up-to-date definition further down in the file.

Author: Liamolucko

WIT.md

@@ -116,43 +116,6 @@ keyword ::= 'use'
           | 'stream'
 ```
 
-### Identifiers
-
-Identifiers are used for the names of functions, parameters, etc. Identifiers
-can either be specified raw as a sequence of characters or as a string literal.
-As a string literal an identifier is allowed to be any valid unicode string,
-including those that might overlap otherwise with keywords. For example an
-identifier can't be `use` but it can be `"use"`:
-
-```wit
-identifier ::= keylike+
-             | string
-
-keylike ::= '-'
-          | 'a' ... 'z'
-          | '0' ... '9'
-```
-
-Strings are intended to be the same format as strings in the WebAssembly text
-format except that they're always valid unicode and don't have raw byte escapes:
-
-```wit
-string ::= '"' stringchar* '"'
-
-stringchar ::= c       if c == \u{9} or (\u{20} <= c <= \u{10ffff} and c != \u{7f}
-             | escape
-
-escape ::= '\\'
-         | '\"'
-         | '\''
-         | '\t'
-         | '\n'
-         | '\r'
-```
-
-In subsequent code blocks of lexical definitions, `identifier` may be referred
-to as `id` for short.
-
 ## Top-level items
 
 A `wit` document is a sequence of items specified at the top level. These items