Question from Fletcher: what methodology applies when we identify a conflict between different concepts of time in our network of theories?  How do we decide how to modify our concepts?

K: In Barbour’s account, for example, treat his Leibnizian principles as unshakeable.  How can other concepts be modified to make them work?  But we need to appeal to idea of the “best option or fix” where a full theory of these desiderata is now missing.