When we consult reason, all causes and effects seem equally explicable a priori; nor is it possible to assign either of them, by the mere abstract contemplation of their nature, without consulting experience, or considering what we have found to result from the operation of objects. And if this proposition be true in general, that reason, judging a priori, finds all causes and effects alike explicable; it must appear more so, when we compare the external world of objects with that world of thought, which is represented as its cause. If reason tells us, that the world of objects requries a cause, it must give us the same information concerning the world of thought: And if the one seems to reason to require a cause of any particular kind, the other must require a cause of a like kind. Any proposition, therefore, which we can form concerning the cause of the former, if it be consistent, or intelligible, or necessary, must also appear to reason consistent or intelligible or necessary, when apply'd to the latter, such as you have described it; and vice versa. It is evident, then, that as far as abstract reason can judge, it is perfectly indifferent, whether we rest on the universe of matter or on that of thought; nor do we gain any things by tracing the one into the other.
This passage is scored out. Then Hume changed his mind, and wrote in the margin: "Print these lines, though eraz'd." Then he changed his mind again and scored that out. Then he printed the same sentence in the margin again. Then he scored that out again. The passage substituted for it in the final version was this one:
If reason (I mean abstract reason, derived from enquiries a priori) be not alike mute with regard to all questions concerning cause and effect; this sentence at least it will venture to pronounce, That a mental world or universe of ideas requires a cause as much as does a material world or universe of objects; and if similar in its arrangement must require a similar cause. For what is there in this subject, which should occasion a different conclusion or inference? In an abstract view, they are entirely alike; and no difficulty attends the one supposition, which is not common to both of them.
This wavering in Hume gives a good picture of one sort of limit one always comes up against in History of Philosophy. Because we have some of the revisions, we know that Hume wavered. But we don't know why. Was the reason a matter of language? In other words, did he feel, for whatever reason, that there was something awkward or unclear about the original passage, something that he perhaps didn't quite find definable (hence the wavering)? Was it a matter of dialogical structure, i.e., did he think that the way the first passage framed it introduced an issue he wasn't sure he wanted to bring in? Was it a matter of the argument itself, i.e., did he waver about the cogency of the original, and then finally decide to rewrite it? We can probably make some progress with these questions -- for instance, that the passage eventually substituted is much more concise is perhaps evidence of something. But there will always be a limit. And the same limit is found everywhere, just less noticeably. We can know, to a considerable degree, how passages in philosophical texts fit into their context. But we don't always know much about why this way of saying things was chosen over that way of saying things -- indeed, we don't always know what 'that way of saying things' is at all. The production of a philosophical text is a process of selection, and we rarely get any peek at that. Even when we do, as in the above case, what the peek shows us is less the nature of the selection and more just how much we don't know about the selection. On the other side of the text is a philosophical mind; but we don't actually see anything of that mind except the final selected presentation (and sometimes a few pre-revision selected presentations). What we get is a simulation of the workings of the mind of Hume (or whomever else) on a particular set of points. It is, to be sure, a simulation that comes from the very mind whose thought it is simulating, and that can be quite important. But it's a simulation, a production in which the author does more to act on us than to show us himself (and some philosophical authors go to great lengths not to show themselves much). And I think that's a limit we cannot really overcome in HoP. We can, as I noted above, make some headway on the questions that arise. But there's always that limit.