Historically our theories of truth fall into two broad types: correspondence theories and coherence theories. You seem to be saying that for the relativist a correspondence theory of moral truth will not work for moral principles but that a coherence theory might.

Let me add this. I think that Ann is correct that the word 'knowledge' might mislead people here. It may be better to speak of "having good reasons to accept" principles or "having a justification for" principles.

In any case, what you are saying is that a relativist can develop a coherence theory of justification in spite of the fact that he or she is committed to the view that in some cases more than one conflicting moral judgment is true.

Can you develop this idea of coherence somewhat more?