There's a kind of book I like to read. It is non-fiction, but not the "fun" non-fiction, where the author includes themselves in the story, doing research, doing interviews, contemplating the subject matter, blah, blah, bleeping, blah. Certainly not the wacky non-fiction where the author is the subject, engaging in some kind of year long insanity like trying to reduce their environmental impact to zero or whatever. No, the non-fiction I am referring to here is often called a "monograph". You can recognize a "monograph" in several ways. First, the subject matter is very narrow, and the writing style is formal? Stilted? Certainly scholarly. None of this, and then I got lost trying to find this guy's basement office where he attempts to measure psychic presence stuff. Second, the publisher is a university press, or, conceivably, an scholarly imprint of a trade press. Third, the main text of the book is relatively short (on the order of 200 pages), and endnotes (hopefully at the very end, and not at the end of each chapter) plus things like a bibliography and index constitute a third of the book. There may be a few pictures, but they probably aren't in color. The cover matter (cover art, title fonts, description of the book and author) is restrained and probably doesn't include any review matter along the lines of, gosh, wow, you really MUST read this fantastic book right now it's so compelling. (Altho if a university press could get someone plausible to say that about the book, they would probably put that on the cover. Of course, they have their own sense of who is plausible as a reviewer.)
I love this stuff. I can't read that much of it (I need my trashy, narrative thrusty, ahem, genre fiction in between, along with the cotton candy kind of narrative non fiction), but it's the entree of my reading meals. You _can_ just read an appetizer and then have dessert, but it's not the same.
Some monographs read just fine ignoring the endnotes: I'm familiar enough with what they're talking about that I know what those notes are probably referring to. I'm here for the argument, not the information. Some monographs I _have_ to read the endnotes, because portions of the argument are developed in the endnotes (I disapprove of this). Some monographs _I_ have to read the endnotes, because I'm not familiar with the information and thus can't follow the argument without doing a little side research along the way. The strategy for a parallel read (text and notes) is to have a finger and/or bookmark in the text and in the sources. That's a lot of work, tho, so I prefer when the book supports losing and refinding my place in the notes. Ideally, that's by printing at the top of each and every page in the notes (which are at the END OF THE VOLUME, not buried in arbitrary spots throughout the book) which pages the notes are referring back to. That also lets you read the notes, and go read the text as secondary (which, believe it or not, is a great strategy in some monographs).
As near as I can tell, all major e-reading file types support linkages between the number in the text and the note it is referring to, generally in the form of an anchor tag and target. Unfortunately, as you can probably imagine, manually creating all those tags and targets is a painful and thankless task, so most publishers are not doing it. While it is quite possible to create software that would support creating these linkages in an easier way, it is probably not possible to create software to do it fully automagically (well, actually it probably is, but writing an expert system for these purposes does not sound remunerative or even personally rewarding, even to me, and I doubt you could even communicate the requirements to most software engineers. And then people would persist in complaining about all the weird errors it committed in books with, say, lots of tables or quantitative anything in the text). I'm not holding out any hope that this problem gets fixed soon, as a general rule, certainly not for backlist titles that are autoconverted.
I have been doing some experimenting with bookmarks, but advancing bookmarks in the sources is a pain in the neck in person with fingers; doing it using the bookmarking facilities in kindle and elsewhere is worse. And so, I have resorted to Dual Screening.
I started by looking at the book in Adobe Digital Editions. It is pretty, but longform reading on an active screen is really tough for me. I can do it, but it gives me a headache, and the more I do it, the worse that headache gets until it is a migraine. No fun. When I got the kindle edition, that was better, and I used Adobe Digital Editions to be a bookmark in the sources. That was nice, but awkward. It works okay sitting in my chair, but not lying in the bed. Once I got Bluefire up and running and the kids were no longer competing for the iPads, I had the chance to experiment with the Sources in Bluefire and the text on the kindle. This worked great, altho I personally would probably prefer two kindles. In this particular instance, that would require me asking the publisher to resend the kindle text to a second kindle (since it was going via the email link -- it's a Netgalley thing), which seemed like more hassle than it was worth.
Here's the executive summary for anyone who has read this far:
(1) There's nothing about this that scholars can't adapt to. The laptop + kindle solution also supports note taking and side research really effectively. I expect this one to be the default, altho it does trap you back at a surface or at least in a chair.
(2) E-editions which make you pay for every copy SUCK. Having figured out how to do this, any scholarly text that didn't let me have enough copies to at least dual screen would be a scholarly text I would avoid.
(3) Endnote linkages matter a lot less than I thought they would.
(4) I can suddenly see a purpose for headers (altho not footers) for chunks of a file. Each section of the notes, say, could be marked off by chapter, with a header that indicated the chapter. This would allow reflow, but let a scholar "flip" through the notes to the spot they wanted much faster than looking painstakingly for the breaks with the headings.
(5) None of this addresses the basic problem of citation caused by reflow and the lack of stable pagination that is currently pushing the academic world towards pdfs. I don't, personally, see the problem: cite the kindle edition and the location number and call it good. Page numbers aren't (necessarily) stable from hardcover to paperback or across editions anyway (altho to be fair, they usually are stable on the hc/pb pair of a monograph from a university press).