More on relative truth and disagreement. Let 'Pa' denote that P is true relative to Anne. I suggested before that such relativity claims are absolute: even Bob should agree that 'abortion is wrong' (P) is true for Anne. It's just that it is false for him, and thus false simpliciter (from his context of assessment). So he would grant Pa but not P.
Jack now brings to my attention the following biconditional: (P iff Pa)a. We've seen that Bob will deny (P iff Pa). But he must also recognize that P would be assessed differently relative to Anne, such that the biconditional will come out true for her.
Now here's a plausible principle: as responsible epistemic agents pursuing knowledge, we should not want to lead our fellow inquirers into false beliefs. So Bob should be reluctant to convince Anne to deny P after all. Why? Because, by the Anne-relative biconditional, she should then also deny Pa, but Pa is an absolute truth. (I'm assuming that her context of assessment won't change when she changes her belief about P.) Indeed, the point may be made even more directly by noting that, in virtue of being true-for-her, P is surely the appropriate belief for Anne to have, the one she should have, even if Bob would judge it technically 'false'.
So if Bob can grant that Anne is believing appropriately, is there really any genuine disagreement left? Maybe it is just the kind of non-cognitive opposition you get with emotivism -- merely a matter of cheering for two different teams. You need not think the other person has made any mistake, but you are oppositely aligned and so destined to clash. But I guess even the meaning relativist can tell this kind of story, so it's no longer clear what distinctive benefits are offered by truth relativism. (Or should the truth relativist hold even 'appropriate belief' to be a relative matter? That seems to be going too far...)