Now it is a natural consequence resulting from the experience we have of the value of truth amidst the transactions of life, that mankind will speak the truth in all cases, when it appears useful and accords with their interest to do so; as well as that in all other cases where the contrary consequences appear, men will be strongly tempted to falsehood; being only prevented from using it by observing that a superior value is contained in observing a general rule prescribing truth indifferently, whether for or against their interest. It thence follows as an axiom, that we place dependance on the veracity of men, in all cases were we cannot distinctly perceive any motive to falsehood; and in like manner that we proportion our jealousy of the truth of their assertions, according as we may suppose them influenced by any circumstance of self-interest.
Lady Mary Shepherd, Essay VIII from Lady Mary Shepherd, Essays on the Perception of an External Universe, pp. 325-345. Like Hume, Shepherd recognizes the problem of testimonial evidence as a causal one, and, also like Hume, is applying her account of causation to it. Since they have very nearly diametrically opposite accounts of causation, it's unsurprising that they get very different results when considering what we can know on the basis of someone's say-so. One of the interesting things here is that this is an abstract argument for the rationality of what is historically (and misleadingly) called the principle of credulity: that we assume people speak truth unless we have some reasons to think otherwise -- most defenses of the principle of credulity in this period are empirical rather than abstract, arguing that this is in fact what human beings will inevitably do, regardless of whether anyone considers it rational or not.
