I like your cynical model and would emphasize it more. I.e.:
Step 0, the listener has a preexisting opinion of Sarah. Step 1, the listener hears that Sarah did Thing X. Step 2: If the listener had already liked Sarah, then they’ll rationalize why Thing X was actually fine under the circumstances; or if the listener had already disliked Sarah, then they’ll rationalize why Thing X sucks and Sarah definitely sucks. (Or if the listener was neutral about Sarah, they’ll probably pick a side based on other vibes, e.g. how they feel about you the speaker.)
I think it’s worth thinking about conversational implicatures here. In general, in everyday speech, people talk about other people much more often than people talk about norms and ideas (present company excluded). So in everyday speech, if Alice say “Sarah did Thing X which is bad”, then it’s usually a correct assumption that the main takeaway / upshot / purpose of this speech act is that Alice is trying to communicate the message “we should all think less of Sarah”. And then the listeners will think to themselves: “Do I basically agree or basically disagree with this main message that Alice is trying to communicate?” And if they basically disagree with that main message, then the detail “Thing X is bad” will get ignored at best and sign-flipped at worst.
I like your cynical model and would emphasize it more. I.e.:
Step 0, the listener has a preexisting opinion of Sarah. Step 1, the listener hears that Sarah did Thing X. Step 2: If the listener had already liked Sarah, then they’ll rationalize why Thing X was actually fine under the circumstances; or if the listener had already disliked Sarah, then they’ll rationalize why Thing X sucks and Sarah definitely sucks. (Or if the listener was neutral about Sarah, they’ll probably pick a side based on other vibes, e.g. how they feel about you the speaker.)
I think it’s worth thinking about conversational implicatures here. In general, in everyday speech, people talk about other people much more often than people talk about norms and ideas (present company excluded). So in everyday speech, if Alice say “Sarah did Thing X which is bad”, then it’s usually a correct assumption that the main takeaway / upshot / purpose of this speech act is that Alice is trying to communicate the message “we should all think less of Sarah”. And then the listeners will think to themselves: “Do I basically agree or basically disagree with this main message that Alice is trying to communicate?” And if they basically disagree with that main message, then the detail “Thing X is bad” will get ignored at best and sign-flipped at worst.