With regard to the AHA survey, I thought it apparent that it regarded trust in historical interpretations presented by these sources (teachers, non-fiction books, etc.).

In response to one of your later comments on Foucault, he wrote in the introduction to the Archaeology that discourse and its relation to the author are non-transcendental. In the conclusion (where he answers many of his critics), he later argues that no person's perspective can outlive them. In the first chapters of the book, Foucault eventually acknowledges that the interpretation of the author sometimes does not outlive the very moment within which it is created. As all written discourses suffer this problem, the issue becomes that we cannot approximate the beliefs the author even held. Additionally, the structures in which the discourse originated are fundamentally different to those in which we today are interpreting them.

Therefore we cannot know what the author intended for us to receive in interpreting their writing, much less understand the message of the past without its ceasing presentness. Epistemically, the problem is that we receive the document from a conscious process of selection (rules of discourse) that inevitably distorts, rather than reveals.