Extensions to other times and other objects

A question I raised in an earlier Hume post went as follows:

/>/One of Hume's central doctrines is that we cannot discover causal relations via impressions; that's clearly what his range of examples is supposed to show (bread, billiard balls, etc). But in part 2 of section 4, Hume glosses his doctrine in a slightly different way: "As to past Experience, it can be allowed to give direct and certain information of those precise objects only, and that precise period of time, which fell under its cognizance: but why this experience should be extended to future times, and to other objects...." The problem here seems to be not about causal relations, but about our knowledge of object and other, as well as our knowledge of past and future times. So an intriguing question: is causal information non-discoverable because object-other informationis non-discoverable? Or perhaps because past-future information is non-discoverable? (Which thing can't we know via impression that leads us to miss out on natural patterns?)/->/

Let me see if I can make my central idea a bit clearer. (1) I wonder, given Hume's remark here quoted, why he thinks we cannot tell how to extend the information we have about a particular object to some other object. Similarly, (2) I wonder why he thinks we cannot tell how to extend the information we have for some period of time to some other period of time. 

Questions (1) and (2) seem different to me than the stock Hume doctrine that we cannot have causal information given that we cannot have an impression that corresponds to the necessity that effects follow from causes. The stock doctrine is plainer to accept, it seems, since necessity is logical information, and on the Humean picture logic belongs among relations of ideas rather than matters of fact. But is the this-ness of an object versus the that-ness of some other object logical in that way? Less idiosyncratic: is the individuation of objects logical for Hume? The time version of my worry (in (2)) is similar, if future times are simply regarded as times other than the past or present ones. Is the this-ness (perhaps the now-ness, or the past-ness) of a time versus the that-ness (then-ness) of some other time a logical feature of it?

There's a strong Cartesian reading we could give which would treat pattern-y features of our precepts as logical; but no self-respecting empiricist could possibly go there, I would think. To see things that way would completely hollow out empiricism, since sense information would be little more than a logical edifice built around some kind of raw stimulus. 

Maybe Hume has something in mind about abstract categories of thought that are inaccessible to sense, and that capture some of the features I'm puzzling over. Maybe his treatment of space and time (in the Treatise) would shed more light on this. 

That's where I'll look next.