Abstract
It is thought that a valid inference to a logically necessary conclusion must proceed from entirely necessary premises. Counter-examples show this is false. Perhaps while the truth of a necessary proposition may follow from non-necessary premises, its necessity cannot so follow. Counter-examples show this to be mistaken. Must anyone who comes to know the non-necessary premises employed in the various counter-examples have prior knowledge of the necessity of the conclusions of the counter-examples? I argue against this. It is true that, for any necessary proposition, there must be necessary premises from which it may validly be inferred; but no one need use these, or know these, or know how to use them, in order to know the necessity of any proposition.