Jakob Nielsen's Alertbox for July 26, 1998:

Electronic Books - A Bad Idea

Two new electronic book products have gathered extensive press coverage, including articles in ABC News, New York Times (access requires registration), San Jose Mercury News, and WIRED.

Both products are really tablet computers, though this phrase is less sexy than the metaphoric term "electronic books". In particular, the SoftBook has nice leather binding, going far beyond the traditional (and boring) industrial design used for other portable computers. The RocketBook has a more traditional industrial design, typical of leading electronic gadgets.

Being a firm believer in tablet computers for many tasks not involving heavy data entry, I applaud two new tablet designs. We don't need every computer to have the same user interface or to come with a keyboard. Unfortunately, neither website mentions their product's screen resolution (dots per inch). Low-resolution monitors (including all computer screens until now) have poor readability: people read about 25% slower from computer screens than from printed paper. Experimental 300 dpi displays (costing $30,000) have been measured to have the same reading speed as print, so we will get better screens in the future. People will simply not read long texts at a reduced reading speed, so unless they have much better screens, electronic books will have a problem.

Another project getting much attention is "the last book" project at the MIT Media Lab. This project aims at producing a computer in the form of a folio: a bound set of pages using "digital ink". Users would move forward in the information by flipping pages just as they do in a printed book. However, the electronic book uses special paper that can be rewritten under computer control. Thus, a single physical book can turn into any desired work simply by downloading the corresponding file. Hopefully, the MIT scientists will eventually invent a way to make the "digital ink" display characters at a high enough resolution to gain the same reading speed as print.

Even when electronic books gain the same reading speed as print, they will still be a bad idea. Electronic text should not mimic the old medium and its linear ways. Page turning remains a bad interface, even when it can be done more conveniently than by clicking the mouse on a "next page" button. It is an insufficient goal to make computerized text as fast as print: we need to improve on the past, not simply match it.

The basic problem is that the book is too strong a metaphor: it tends to lead designers and writers astray. Electronic text should be based on interaction, hypertext linking, navigation, search, and connections to online services and continuous updates. These new-media capabilities allow for much more powerful user experiences than a linear flow of text. Linear text may have ruled the world since the Egyptians learned to produce arbitrarily long scrolls of papyrus, but it's time to end this tradition. Nobody has time to read long reports any more: information must be dynamic and under direct control of the reader, not the author.

Some Electronic Books That Work

Two types of electronic books do make sense: In both cases, the key point is that the "electronic books" are not intended to be read on a screen: they are traditional paper books and linear audio readings, respectively, and are simply manufactured and distributed in a more efficient manner by using the Internet.


See Also: