When reading physical books (particularly during Inspectional reading), a skilled reader naturally maintains their reading position in multiple sections of the book simultaneously. For example, this might be a “stack” operation (keeping their place while referencing another), a “dog-earing” operation (accumulating a set of places to focus on; see Askwall), or a parallel reading operation (comparing several passages). Digital readers make this kind of operation very difficult.
Because screens are often bigger than physical books, one might use multiple windows to manage this, but Parallel reading is mostly impossible in digital reading. Even on desktop operating systems, most digital reading applications won’t even let you open a second window viewing the same document. This limitation is exacerbated by the problems described in Poor performance disrupts nonlinear reading in digital reading.
As an alternative, one might use bookmarks or an explicit in-system structure to manage reading locations. But digital readers’ bookmarks—when they have them—are always too heavy.
LiquidText is the one exception here: it uses multitouch to create ad-hoc bookmarking and provides a canvas where one can accumulate increasingly durable references.
The Amazon Kindle has a simple one-direction navigation stack for a limited set of operations, like jumping to a footnote or to a place in the table of contents. This helps, but only in a small subset of cases.
Askwall, S. (1985). Computer supported reading vs reading text on paper: A comparison of two reading situations. International Journal of Man-Machine Studies, 22(4), 425–439. ~https://doi.org/10.1016/S0020-7373(85)80048-1~