The Information Design of Footnotes

🌿 Draft — first pass, from a long-standing note. Rough but the argument is complete.

Why are books still so badly designed for the apparatus that supports their argument — the footnotes, the endnotes, the bibliography?

Consider the standard arrangement. Footnotes sit at the bottom of the page, forcing the eye down and back up, when they could sit in the margin alongside the text they annotate. Endnotes are worse: banished to the back of the book, so following one means a two-handed excavation and a lost place. The bibliography is further back still, and the citations that point to it are terse to the point of uselessness — a name and a year, rarely a title, so you cannot tell whether the thing being cited is a book you know, a paper you should read, or a passing aside.

Here is a concrete case, from Philippe Aigrain's Sharing Culture:

In 1997, we published an article in First Monday… Attention, Media, Value and Economics (Aigrain 1997) was a contribution to a debate on the "attention economy" launched a few months earlier by (Goldhaber 1997) and (Ghosh 1997). Based on earlier work (Aigrain 1996), the paper analyzed how much people are willing to pay for one hour of interaction with music or video.

Four citations in three sentences. What was Goldhaber's earlier work? Turn to the back, scan the G's, find it, register it, turn back — except now you have lost the thread, because there was no back-link, no way to return to exactly where you were. Then do it again for Ghosh. Then again for the two Aigrains. By the time you return, the sentence has gone cold.

None of this is hard to fix. Side-notes instead of foot- or end-notes. Citations that carry enough information to be legible in place. Bidirectional links between a reference and the point it supports. On the web this is trivial; in print it has been solvable for decades. Is this really the best we can do with books in 2026?