Summary: Designing information for paper was largely about managing reference distance. On the Web, the reference distance is zero. A completely different set of design requirements apply.
Linear information design
Traditional information design thinking has always been linear. This is a consequence of the medium in which the vast majority of information was presented: paper.
Whether you roll it into a scroll or chop it up into pages and bind them into a book, paper is a linear medium. This presents a fundamental information design problem: some of the things you want to talk about in the real world have complex relationships that cannot be laid out in a single straight line.
Linearizing a complex world
Of course, human beings can only take in information in a straight line. But when a human is trying to understand a complex subject, they are not simply accumulating a sequence of facts. They are slowly building, clarifying, and correcting a multi-dimensional picture of the world. That does not happen through reading alone, but through a combination of reading, conversation, thinking, experience, and experimentation.
Their reading, therefore, while it is strictly speaking linear for each individual reader, does not follow a single predictable course that the writer can anticipate and provide. And because the world is complex, the writer must often mention subjects without fully explaining them at the point they are mentioned. This generally requires some mechanism, implicit or explicit, for the reader to jump to a description of that topic if that is what they need in their current stage of picture building.
These mechanisms may be an explicit link or cross reference, or they may simply consist of explicitly naming the subject. But in either case, the writer encounters the problem of distance.
The problem of reference distance
A book is linear, and therefore any reference to another subject is either a reference backwards or a reference forwards or, if the book does not treat the subject itself, a reference out. In each case, these references have a specific distance attached to them. A reference to the previous paragraph is shorter than one to a previous section or a previous chapter. A reference to an original manuscript in the Vatican archives has a very large distance indeed.
The distance of a reference has a direct impact on its usability. The greater the reference distance, the more work the reader has to do in order to access that subject when they need it. When you organize a book, therefore, one of the things that you spend a lot of time on is minimizing the distance of the most likely references. You plan the outline of the book so that the reader will have to travel the shortest distance to find the next thing they are most likely to want to read.
The biggest manifestation of this, of course, is the overall order of the book. The shortest reference distance is “next”. The narrative line of a book is an attempt to make sure that the most likely course of the reader at any point is to the next sentence, the next paragraph, the next page.
After that, you worry about how far they will have to go for any asides they need to make. This affects the positioning of any pictures or diagrams they might need to refer to, for instance. It also involves inserting capsule definitions and summaries of subjects that are referred to in the text — because the main treatment of those subjects is a long distance off.
As writers we do this kind of thing all the time. If we mention a subject by name in one sentence, we generally use a pronoun for it in the next sentence. But if we then insert a couple of extra sentences after the first sentence, the pronoun in the second sentence is now too far from the thing it references, and we edit the sentence to name the subject explicitly again.
The design of books, therefore, is heavily influenced by the need to minimize reference distance in a linear media.
Hypertext and the zero reference distance
A hypertext media, such as the Web, is not linear. Its references are not “forward”, “backwards”, or “out”. They are just “to”. There is no difference in the distance of references. In hypertext, reference distance is effectively zero.
In this environment, there is no need to spare a moment’s thought to minimizing the average reference distance. The average reference distance is zero. (Actually, within individual pages of a hypertext, there can still be some reference distance. This still requires the same kind of design decision making, on the micro-scale, as in the paper world.)
Information design in a zero reference distance media
Once the average reference distance drops to zero, much of what drove information design in the linear world is simply moot. There is just no need for it.
There is no need to decide which of the several related topics you should treat next. Simply link to all of them.
There is no need to insert capsule definitions of subjects that you will not treat until later — your later is not the reader’s later. Just link to them and let the reader decide which order to look at them in, based on what they need and what they already know.
There is no need to try to determine a hierarchy of topics or to classify one topic as subordinate to another (this is simply a way of reducing reference distance between higher level topics).
There is no need to include material on a tangentially related subject at all. You just link to an appropriate source. (The reason for including it in a paper book was to save the reader from the long reference distance involved in finding a source on paper.)
Rethinking design for zero reference distance
The world of zero reference distance has its own set of information design challenges. For one thing, zero reference distance is not the same thing as zero semantic distance. And establishing the context of a topic when you have no idea where the reader came from presents some interesting design challenges.
Most importantly, perhaps, the reference distance is only zero when you provide good links at all the appropriate points of reference. Otherwise, the reference requires a search, and while a Web search is a shorter distance to most information than a trip to the library, it is often far from zero.
Book design was about optimizing the linear order of a small information space. On the Web, the sheer vastness of the information space threatens to undo all the advantages that the zero reference distance provides. A disciplined link strategy is to Web content what the efficient organization of the TOC was to book content.
But the key thing is, the Web is a radically different environment with a radically different set of design opportunities and challenges. Information design for the Web is fundamentally different from information design for paper. There is no technical solution — no transformation technique — no single sourcing tool — that will make your book-designed content work on the Web.
Every Page is Page One. The reference distance is zero. Design accordingly.