Flew's challenge to the religious believer asks him to specify what would count as a disproof for, e.g. ‘There is a God’. A statement of such a specifiable condition I called an ‘empirical denial’. In my earlier paper I was concerned to show that a statement is a statement whether or not it has such an empirical denial. I was not particularly concerned to show that there are some statements which do not have an empirical denial; my concern was to show that it is not essential that they do. It was in connection with this point that the statement ‘John loves Mary’ was discussed. But this very same example may indeed serve as an example of a statement for which, in certain complex situations, there is no empirical denial. That is, in some circumstances, no matter what empirical conditions are specified, when the statement ‘John loves Mary’ is made (if they are not merely necessary conditions), they could obtain and still John might love Mary, as the developing situation would make clear. If the condition specified is that John is cruel to Mary and shows no concern for her, details could always emerge in such a situation to explain why, even though he loved her, he was cruel and showed no concern. And so on.