Utterances in common language can have drastically different meanings, depending upon the circumstances in which the utterance is made. Sometimes understanding another's speech involves deliberate and careful consideration, particularly when the conversation is about the complex rules and procedures that convention holds in such high regard. Other times such as everyday common parlance require no effort. As a result of both the speaker and the listener being common practitioners, there's little to no further consideration needed.

There is an actual difference between understanding a statement when completely divorced from a speaker and understanding belief statements of a speaker.

Understanding that difference happens by virtue of carefully considering the truth conditions themselves. I take the source of an utterance to be an individual speaker. When we take account of another individual's belief statements, we must err on the side of what they mean as compared to what the truth conditions of the statement could mean if someone else uttered it in an entirely different set of circumstances. Utterances often have several different acceptable uses/meanings.

Understanding an utterance consists of an audience drawing correlations between the same things as the speaker when one of those things is the utterance itself. Successful communication with speech is existentially dependent upon shared meaning; a plurality of capable creatures drawing much the same correlations between the utterances and other things. That's precisely what successful communication with speech always consists of. The correlations drawn by the speaker may or may not be difficult to figure out, understand, and/or interpret... depending upon the circumstances.

"Hey! I said shut that door!", when spoken in the dead of the coldest of winters by a visibly unsettled speaker for the third time has much different meaning than "Hey! I said shut that door!", when spoken by much friendlier eyes and softer tones by one who just wants a bit more privacy. I would have no issue whatsoever with saying that people in either set of circumstances have little to no doubt over what the speaker means, and that such mutual understanding happens in rather autonomous fashion(without deliberative effort).

However, it's still well worth noting the fact that the utterances do not share the same meaning despite the fact that they share the same terms in the same order. When we separate the statement from the circumstances in which an utterance is made, we have cleaved off irrevocable and crucial aspects of the meaning. Many of the correlations drawn(much of the meaning of an utterance) include the circumstances in which the utterance is made. Understanding the speaker requires considerations including the particular circumstances. Clearly.

So, divorcing utterances from the speaker and the overall context/circumstances in which the utterance is made results in inadequate ground for any conclusions about the speaker's meaning when there are differences between that and conventional and/or literal(previously accepted) meaning(s). Sometimes there are.

A more complex example of the same statement being a part of two very different sets of circumstance, which does require some rather deliberative effort to understand would be what's commonly called the "Gettier problem". "The man with ten coins in his pocket will get the job" is a prima facie example.

Gettier wants us to believe that Smith arrived at that belief statement by virtue of using and/or following the rules of logical entailment. Smith believed that he would get the job and that he had ten coins in his pocket. Smith was thinking about himself, and for all intents and purposes, that belief was well grounded. I mean, he had heard it from the individual in charge of doing the hiring. He was a shoe-in. We all grant that much.

No problem.

But what can we say about the subsequent belief(resulting from applying the rules of entailment) that Smith arrived at? It was Smith's belief being taken into account. We already know that Smith believed that he was going to get the job. So, we also know that when Smith inferred that the man with ten coins in his pocket would get the job, Smith was still talking about and thus was referring to himself. He did not get the job. Smith's belief was false. We cannot possibly understand that unless we think about the differences between the truth conditions of Smith's belief as compared/contrasted to a literal interpretation of the same sentence(the statement completely divorced from the speaker).

So, I remain quite convinced that truth conditions are irrevocably entwined with meaning, and that keeping those in mind is quite often a crucial - and often deliberate - step to understanding an utterance in a language you already know.

yes, it's interesting article by Davidson, in which Davidson overemphasized the need to adapt theory to a particular set of circumstances. To get a passing theory you need a prior theory, and all Truth and Interpretations is about the latter. The text should be read together with the section of Truth and Interpretation that deals with the problem of convention and illocutionary force, and metaphor (on which the realm of semantics has almost nothing to say, i agree with that). Rorty was inclined to see in this Davidson/Derrida, which is of course a preposterous description, but it's the result of Davidson's own fault. I think Davidson has a valid point there in the sense that, to language, there is a part that is not couched in terms of general theory, or rules. but in terms of idiosyncratic features, linguistic individualism of sorts. Barbara Johnstone published a great book on that with no reference to Davidson. I think Otto Jespersen somewhere stated that language combines elements of the general and the particular (individual style etc.). Interestingly, in a conversation with Quine Davidson said that we interpret not language, but persons. Again, an overemphasis, but it gives you some clue about what he intended to say. Frankly, it's extremely difficult to reduce, for instance, the process of (legal or textual) interpretation to a process of application of rules. In that I agree with Davidson. However, I do not think that the process is erratic or subjective or not bound by standard epistemological criteria such as truth, relevance, simplicity and coherence.