Reference vs. Interpretation
In Brains in a vat Putnam argues that if an ant walks around and the path it traces happens to “draw” or “resemble” Churchill that doesn’t imply the image left in the sand represents Churchill, because the ant had no way of knowing about Churchill. The same about a description of trees found in a text that happens to be written by monkeys randomly typing on a typewriter. The words there do not refer to anything, says Putnam. So from a reference perspective intentio auctoris matters. (see Reason, Truth and History by Putnam, Brains in a Vat).
Now about interpretation. Let’s say there’s a culture for which a volcano is a god that communicates to the villagers about its mood, when the volcano is angry, there’s an eruption and so on. So the villagers interpret different things about the world and react accordingly, based on how the volcano behaves, they provide sacrifices and so on. Clearly as per Putnam’s argument, the volcano cannot be referring to anger or any other kinds of moods, or cannot be demanding sacrifices, but from a semiotic perspective that is completely different, almost irrelevant. Semiotics is interested in the code created by the villagers that assign meaning to the various states of the volcano, and what kind of interpretation the villagers do based on that code. For semiotics it is irrelevant if the interpretations are “true”, or can be “scientifically proved”, for semiotics what matters is that there’s a signification system where those signs are interpreted as such by a culture. Here intentio lectoris matters more than intentio auctoris. (see The Limits of Interpretation by Eco, Intentio Lectoris: The State of the Art).