A sentence, such as a Godel string, can be given any intepretation whatsoever, so in a sense, when communicating, we have to be benevolent about the interpretation function that we infer our co-communicators are using.
I am looking for a formal treatment of the notion of interpretation functions and the fact that there are multiple interpretations possible for any given sentence and formal system.
I believe this is glossed over in Godel's (1931) Incompleteness Proofs and I have found the following two quotes:
"No sentence can be made both true and false by the same interpretation, but it is possible that the truth value of the same sentence can be different under different interpretations." 
"Any given formal language can be paired with any of a number of competing semantic rule sets."