These are the slides from our presentation at OTESSA 2022 yesterday, describing our somewhat delayed project to try to make the content of e-texts more memorable by inserting artificial landmarks.

We are building an e-reading system to insert/overlay/underlay patterns, glyphs, or other visual (and, in future, audible) landmarks based on a hash of the content of each paragraph, thus trying to achieve an effect not unlike that of random minor blemishes and imperfections on a printed page but (because they are attached to content, not physical location) that remain consistent when the text is reflowed and reformatted.

There is some evidence that such landmarks make it easier to locate previously-read content, and qualitative studies suggest that lack of landmarks may be why e-texts are usually less memorable than p-texts, but we are not aware of anyone else attempting to incorporate such landmarks into an e-reader with the intent of improving recall until now. Our hypothesis is that such landmarks may improve recall of the text by anchoring the abstract words to something more concrete.

This little maggot has been lurking in my brain for a while - I first wrote about it in 2012!