There is a distinction between the writings of a human author and the beliefs of that author. In priniciple, the writings of a particular human are shaped by beliefs of that person. There might be cases in which that is not so, but it is at least ordinarily the case that human writings are, to one degree or other, shaped by beliefs of their authors.

Authorial beliefs ordinarly play a causal role in the process of human writing. Someone may write x, in part, because she believes that x is true or, for that matter, false, funny, hurtful, profound or silly. More basic still, someone may write down x, in part, because she believes she is able to do so. Whether or not all the beliefs of an author play a role in the writing process is for me less clear. There is reason to believe that no aspect of the author is dispensible in the causal process that gives rise to a text. If so, even authorial beliefs that are about very different matters can be said to play a role in the writing process. But let us leave this issue to the side.

That human writings are ordinarily to some degree shaped by at least some beliefs of their authors does not mean that a full picture or inventory of their beliefs can be easily constructed from their writings, even when those have been preserved. Often features of the text are explainable and even plausibly explainable in terms of differening beliefs. A human author having written x might be the partial result of having belief y, but might be plausibly explained in terms of her having belief z.

Furthermore, the beliefs of people on certain topics can change quite radically over time. Just because a human author believed certain things when she wrote a particular book, paragraph or sentence does not mean that she believed them ten years later. That being said, there is also a lot of continuity and stability when it comes to human beliefs. If we have strong evidence an author believed y at t1 then this also counts as evidence, and in the absence of strong counter-evidence, arguably as strong evidence that she believed y at t2.