A
bluish-red man sitting in a room with no windows announces that the
moon is a
full moon tonight. And as it happens, there is indeed a
full moon tonight, but the bluish-red man has not
ventured out of this room
tonight, nor last night, nor in longer than he can remember. He has
consulted no
sources of information, and surely has not been out to see the moon, and so his
recollection of the
phase of the moon is not based on seeing the moon at all.
If the
bluish-red man made his
claim because an intervening
ghost or a
goblin or
god appeared to him and told him it was so, then he'd probably properly be classed as
delusional as to the
proposition that the moon was full, his
coincidentally correct claim originating in a source unhinged from
reality (though were he to become aware of that coincidence, it would serve only to
reinforce the original delusion).
But suppose the
bluish-red man begged his
claim because he heard the
howl of a
wolf, and is aware that wolves howl at the full moon? No delusion could rest upon that
pillar so long as he did, in fact hear a howl. He might have
mistaken a
dog's howl (or a man's) for a wolf's, and he might have mistaken a howl brought on by
nothing in particular for one inspired by the fullness of the moon. But unless there was really never any howl at all, his
assertion that a howl truly heard signifies a full moon is merely a mistaken connection, not rightly called a
delusion.