taking speech from models of characters and tokens
taking speech from models of characters and tokens