I have a kinda like, but mostly hate, relationship with ebook readers. Even though they give us terrific access to huge number of books without taking up shelf space, that is one of the only advantages of standard ebooks. The reading experience is mediocre, interfaces are lackluster, and the primitive layouts are terrible for non-linear content (such as annotations and asides). We — that is to say writers and designers — need to do better. Designers we need to think not just more creatively, but more deeply about what it means to read. Writers need to also think more deeply about how we want our audiences to engage with our ideas, it is ridiculous to carefully craft the prose, but give little care to how those ideas are presented on the page.

In digital, search is easy but finding can be difficult. When I am searching for a particular passage in a print book, it is easy for me to flip to roughly the correct section of what I am looking for. But it is much more difficult in digital. Rarely do we remember exact phrases or words, which means basic text search is less fruitful. Here is the problem: the resolution in digital is still much lower than physical mediums. We rarely remember the chapter something is in, but we remember about what part of the book. What part of the page, the shape and weight balance between the front and back. Whether or not the book laid flat, or if the front or back covers kept folding over because the spine was not broken in. Perhaps I remember underlying and lighting a passage. Or I made some notes in the margins. Perhaps there was a earmark crease from a previous reader. Whether there was a funny footnote at the bottom of the page.

Others call this serendipity, but I think it can be explained in more concrete terms: Memory and thought are contextual and topographical. Our memories are not just audio recorders or stenographers. They are far more plastic and far richer. They are tied to the environment around us, around what we are reading. Our minds are built for a physical world, we create mental maps of the topology of the text¹. We do not look at the page numbers and figure out how many pages we have left, we feel the thickness of the book. We do not count how many lines or inches down a page certain passages are, we look at the page and remember the shape and position, the same way we remember where the teacup is without having to look at it. The same way a chess grandmaster remembers the positions of a board at the twenty-third move. Much of this cannot yet be duplicated in digital, yet ebook readers keep trying to follow the old trope. Designing uselessly skeuomorphic interfaces without reason. And as long as digital books keep trying to play the same game as print books, they will lose. Digital books need to take advantage of their medium instead of superficially replicating old ones.

Instead, we need to provide a new map for our digital books by providing context in other ways. With The Expedition of Humphry Clinker, I offered two separate visualizations, the table of contents that appeared at the end of each chapter, and the explore mode infographic. The table of contents offered a simple way for readers to see their progress through the full story, as letters were opened up and read. But also it gave a little physical map of the chronology was (e.g. the London letters started somewhere on the third line). But this sort of chronology would not work with most texts. Each story is different, each book is different. Our attempts to shoehorn all texts into the same format is ill-conceived and does not do them justice.

The future of the book: just like its past

The future of the book is diverse. Print is not going away, not as long as electricity is scarce and ebook readers expensive. Ebook formats are as diverse as the devices using them. But the goal of a book has stayed the same: communicate ideas from the authors to the audience. That will never change. The goal of the ebook should not be to duplicate the same experience as a print book. That is impossible. There are thing that print does better than digital, and there are things digital does better than print, that is where we should focus our energies.

A few publishers have made terrific improvements over traditional books such as Touch Press and The Waste Land iPad App which not only offers the text, but also scans of T.S. Elliot’s original manuscripts². But since it is only an iPad app, it isn’t readable anywhere else. Kno has been building beautiful interactive textbooks and selling (or renting) them at very reasonable prices. Others such as Pressbooks are also building APIs for their books by default, letting people interact with the texts in their own ways. But these examples are the rare exceptions to a world where the average digital book is at best a giant block of scrolling text.

The regression in digital layout design is not limited to books. It is endemic of every source of digital text. News sources demand using pages to separate texts and pictures galleries, requiring long load times between connected content, breaking reader concentration in the middle of the narrative. Usually doing so to serve more advertisements, and thus generate revenue at the cost of the reader’s cognitive continuity and author voice. But the cost to a page load on a website is much higher than a page flip in a newspaper, magazine, or book. In digital, in the best cases a new page load requires hundreds of milliseconds to render, then more time for the reader to cognize the new landscape. Or worse, many sites now require readers to sit through a 10 to 15 (and sometimes even 30) second commercial, which is akin to dumping a bucket of water on the audience between “Thou shalt break them” and “Hallelujah” in the Messiah. We know that people are incapable of multitasking, yet we demand our readers to do so, and then wonder why they are so stupid. Why they are so uninformed, and never remember anything we tell them. (This one of the reasons I only require page loads between chapters in Humphry Clinker, which are natural breakpoints.)

This is not entirely the fault of publishers and news organizations. After all, we designers and developers are the ones building these things, yet few of us have come up with good ideas for making the digital reading experience better. Yes, in theory the formats we create such as epub 3 allow for interactivity through javascript and the like. But that is not enough. For instance, in digital academic publishing, we are no longer limited by the printed page and the spine of the journal: why are we not required authors to provide their data? Why are graphics and other data still presented as static tables and charts? They were not created that way. The recent fiasco the Reinhart/Rogoff data³ shows the depth of our problem in academic communication. It is shameful that it took 3 years for someone to go back and look at the original data and analysis to discover this fundamental flaw. It is imperative that digital academic books and journals include the data. And in digital, it is easy release the data with an API and allow cross-origin resource sharing for others to test and confirm findings or reuse for new discoveries.

But that is not all, we designers of digital academic literature are even getting the basic fundamentals of book typography wrong: footnotes are actually endnotes, which, as F.L. Lucas says in Style

[Permalink]

Instead of putting notes at the foot of pages, it jumbles them in a vast dump at the back of the book. No normal reader much enjoys perusing a volume in two places at once; further, though he may find his way if he has the patients, from the text to note 345, he may have a tedious search to find his way from note 345 to the relevant passage of the text⁴.

While bad enough in a blog post, where the best solution most people use are endnotes with return links⁵ (which is also is used on sites such as Wikipedia), these are absolutely abysmal in books. As I said above, when we read, we build a topography of the page. In print, the topography does not change, the top of the page is always the top of the viewport. When we track between the citation marker and the note, their position on the page and to each other do not change, so our mind does not need to remap the canvas. But in digital, the canvas keeps moving as pieces flow in and out of the viewport, it is like trying to read a giant map with a tiny magnifying glass, our mind has no way of keeping an accurate view of where everything is. This makes digital publishing for many non-fiction books damn near useless, where footnotes can be integral to understanding the author’s message⁶.

Each book is different, therefore our reading tools should reflect that. As I said when announcing Humphry Clinker, the goal “is not to create the best digital book reading experience, but merely to create the best experience for this book.”

The importance and tone of the aside should shape how they are presented. While some may be able to do with endnotes (to the reader’s detriment), others cannot. The comedic flow and timing of the footnotes in Douglas Adams’ Hitchhiker’s Guide to the Galaxy would be completely destroyed if they required clicking on a link to an endnote or pop up. The notes in Kant’s first critique (which are critical to understanding his philosophy) are lost readers trying to track back and forth between an endnote and the sentence they were just reading.

Yes, many footnotes are abysmal, meander uselessly for pages, and are often full of name dropping crap, but it doesn’t matter. The author included them, the editor kept them, so it is our job as designers, developers, and software engineers to make sure the readers can access them comfortably and with as little distraction as possible.

We have failed at our job.

[Permalink]

The footnote jousting could soon be moot, as the ebook may inadvertently be driving footnotes to extinction. The ebook hasn’t killed the book; instead, it’s killing the “page.” Today’s ereaders scroll text continuously, eliminating the single preformed page, along with any text defined by being on its bottom. A spokesman for the Kindle assured me that it is at the discretion of the publisher how to treat footnotes. Most are demoted to hyperlinked endnotes or, worst of all, unlinked endnotes that require scrolling through the e-reader to access. Few of these will be read, to be sure⁷.

We can not let this happen. Even if we remove asides from the book, we cannot remove citations. Not in this age where unsubstantiated claims run rampant even from the most trusted sources. The digital books we are building today are falling down at the task. They are supposed to convey the author’s ideas to the readers, that has always been the task of the book. But they’re failing at this fundamental task. Our ebooks need to do better. Our ebook readers need to do better. We need to do better.