One back. Two forward?

So having rashly ventured an unusual opinion on Hume after reading only one of his books, I had to go back and read the Stanford Encyclopedia article on Hume by William Morris in order to quickly get up to speed on the rest. And I should qualify my comments from Monday. In the first place, "to show that reason has limits" is not "his whole project," but only half of it, the "negative" or "critical" side according to Morris. Hume also has a constructive intent, which is to provide an account of human nature. Secondly, Hume is obviously very antagonistic toward religion. Now, from reading the Enquiry it seems like Hume might distinguish religion and faith, but any comment should really be reserved until after also reading at least the Dialogues concerning Natural Religion and The Natural History of Religion.

On the other hand, I found the Hume society! Cheap rates, a biannual journal with online archives, and (not so cheap) conferences in Iceland. Hmm ...

Now, Kant. Wow. What a difference from Hume! His style and attitude remind me of Descartes, but I also wonder how much of that is due to reading him in translation. As far as content, I didn't expect him to be so directly engaged with Hume. That's interesting. But having gotten so comfortable in Hume's system, I find myself having to really relax my mind and stretch to connect with Kant. But I think I may have made the connection this morning: universals. Kant speaks as if there are actual concrete things outside of ourselves that we can have a "pure" interaction with. I was making sense of it this morning in terms of realm. There are objects, perhaps Kant is saying, that exist solely in the mental realm. "[P]ure mathematics and pure physics" are the two he cites (p. 17 in my edition). Logic, presumably, is part of that. (Hume refers to these as "relations of ideas" as opposed to "matters of fact.")

Here's the insight this morning: purely mental objects scale cleanly.

Recall that we consciously experience physical objects only via mental objects that refer to them. Reference—the leap from physical to mental—is a lossy abstraction: an abstraction in which information is lost, and the loss is only magnified in further abstractions built upon that reference. The information loss is worth it for the sake of the cool things we can do at higher levels of abstraction. Now, it seems that mental objects without a referent in the physical realm are not subject to this loss of information. There is no physical referent for "7", or "equals," and it appears that "7+5=12" is a clean abstraction, or an abstraction in which no information is lost.

This may resolve two tensions that have nagged at me since I first articulated scale-and-realm. The first is the possibility of clean abstraction (I think of it like vector and bitmap). The second is the relation between scale and realm. There are such things as purely mental objects, and they scale cleanly. But mental objects with physical referents do not.

We'll see if this synthesis survives a full reading of Kant. Feed back to Chad Whitacre.