Conversation

bibliography management is a dinosaur (bibtex, endnote) designed for dead-tree serialized documents... with online use bolted on as a bad afterthought (see footnotes in online docs for eg). Need an online-native model. The serialized print doc should be the afterthought.
4
62
the problem online is that hyperlinks are not stable enough (target could vanish, go obsolete, get replaced), but the solution isn't to retreat to bad ports of dead-tree models... you need things inspired by transclusion (pull in a copy of enough of the cited ref)
2
10
but you still do need to make a distinction between core text, and referential periphery matter. But it should exist as a sort of backdrop/margins fringe. Like if you wanted to preserve an impression in concrete, you'd carve out a slab around the impression.
1
4
ie there's still a first class and second class citizen divide -- your content vs. referenced content. The flattened hyperlink model kinda dissolves that divide and then the core document decoheres. You don't want that. Unless you do...which you sometimes do
Replying to
actually, the idea of a "naked" reference with a very thin pointer as in "John Doe [1] argued that Vulcans are superior to Romulans because XYZ" is a product of scarcity because your private notes about [1] are too hard to expose in a useful way in a traditional research mode.
2
3
Today, you'd want a fat/wrapped reference. The core text would be the same, but [1] would point to YOUR public note about the original text. It's like showing your source code. Show your "reading" of your source texts too.
5
10
If it's a lazy ref where you didn't actually read it, but are citing it for political/social reasons (for eg. you expect the author to be a likely peer reviewer and want to flatter their conceit up front)... that's actually a social problem with peer review based production.
2
3
Interestingly non-academic quasi-scholarly writing does not have this particular problem, though it has others. No kiss-ass citations. I hated those back when I was doing academic writing. I'm working on a whitepaper project now and it's refreshingly free of such imperatives.
1
8
Replying to
Taking a lesson from programming: one of the most important system architecture tools is distinguishing between Model, View, Controller. (And sometimes, View Model.) For texts, “controller” is not apropos, but the work of decoherence is generally about separating View & Model,
1
2
Replying to and
so that one can innovate and build upon prior Models. Different texts also come in at different levels. Some are merely new views on old models; others are surveys of views; and very few are genuinely new models. Narrative coherence generally serves to form a View.
1