[pruss]

One reason not to tie things to screen numbers is that paging by screenfuls, rather than more flexible scrolling, is itself a left-over from dead-tree technology. :-)

Another complication is that not all books are best divided in the same way. A lot of non-fiction books have logical subdivisions worked out by the author. For instance, my book on the Principle of Sufficient Reason is divided into three parts, each of which is divided into several chapters, each of which is divided into several sections, and some of the sections are divided to subsections and subsubsections (e.g., "7.4.2.b"). Logical subdivisions should be preferred to artificial subdivisions.

On the other hand, a number of classic books have their own standard accepted numbering. For instance, the Bible gets cited by book, chapter and verse. Homer is cited by book and line number. Shakespeare gets cited by play, act and scene, with an additional line number as needed. Plato gets cited by Stephanus page numbers (page numbers in Stephanus' 1578 edition) followed by letters that subdivide each page into five sections and often line numbers that subdivide these. Aristotle can be cited by work, book, chapter and section, but is more commonly cited by page number in the 19th century Bekker Greek edition followed by either "a" or "b" to designate the column (it was a two-column edition) and line number. Kant's First Critique gets cited by a pair of page numbers, corresponding to standard German editions of the first and second versions of the text.

There just isn't one a standard way to number things. This makes me think that the ideal would be to have two different location designators. One would be a standardized position whose meaning is relatively independent of the work in question. At least for Western materials, this could be something like section and paragraph numbers for prose (with "section" having different meaning in different works, but always trying to keep to a similar sort of logical or literary division; a chapter for a typical contemporary novel, a "book" for Plato's Republic, etc.), act/scene for plays, section (if available) and line number for poetry, and work-number or title plus the above standard designator for anthologies of the above (so, for an anthology of plays, play title, plus act and scene). These have the merit of being to a great extent a logical and/or literary subdivision rather than merely artificial. While the levels of subdivision would be standardized, and there would be standardized names for the subdivision ("section", "line", "paragraph" in English), the levels could also get "localized" to the work in question. Thus, for Plato's Republic, the "section" could be called "Book" while for a modern novel it could be called "chapter".

And then, in addition to the paragraph numbers, there should be numbering in whatever scheme is either customary for the book in question in the case of older books (e.g., book, chapter and verse for the Bible, Bekker numbers for Aristotle, etc.) or numbering in the author's and/or editor's chosen scheme for contemporary works. A decision how to divide up a work is, after all, a creative division, and as such should be left to the author and/or editor if these are contemporary.

Then it would be up to the reader which, if any, to display in a status line: uniform designators or customary/author/editor designators, and whether to display "out of" information in the case of uniform designators (e.g., "7.14" for chapter 7, paragraph 14, or something like "7/12: 14/128" for chapter 7 of 12, paragraph 14 of 128). The reader would also be able to look up passages by using either uniform or customary/author/editor designators.

Hopefully, with time, more and more works would get cited by uniform designators and dead-tree books would print uniform designators in the page head/foot, though there may always be cases where greater precision is needed than uniform designators provide (like the precision provide by verse numbers in the Bible or Bekker line numbers in Aristotle), and it might be good to enrich the prose uniform designators with a genre-dependent sub-paragraph division.