A HyperTextBook is a reference work that is not meant to be read from beginning to end, but from wherever you start to wherever that leads you. In fact there is no particular reason for a HyperTextBook to have a beginning or end, or a "table of contents" specifying the sequence in which its components are meant to be read, because there is no such intended sequence. It is best approached from a highly sophisticated search engine that might be called a "MetaIndex."

Why do it this way? Because, if you think about it, the whole idea of a reference work is that you suddenly conceive a need to know about something in particular, and so you go look it up.⁵ This is the difference between a HyperTextBook and a "normal" textbook: the latter may have valid pedagogical reasons for leading the student through its contents in a specific sequence, but this intention is lost on the spontaneously curious -- and on those who take responsibility for their own learning.

Most people who visit a Web site are, in fact, driven either by spontaneous curiosity or by some practical need to know a particular thing, which they would like to find in the most efficient and convenient way possible. Hence the HyperTextBook, which is meant to be used more or less as follows:⁶

• "Open" the HyperTextBook at the "Entrance Page" (one way to access the MetaIndex) where you will be offered a choice of browsing schemes -- for instance, "Table of Contents" (a conventional linear pedagogical layout) or "Alphabetical Index" (pretty much what you would find in a conventional textbook, only with more levels, more entries and more links for each) or "Search" (the most sophisticated tool, making full use of the MetaIndex database to construct a list of links ordered by such criteria as "Level of Conceptual Complexity" or "Conceptual Proximity to Current Topic"⁷). Pick one method (probably just the "Alphabetical Index" at this stage) and use it to find the best matches between the HyperTextBook's learning quanta and the subject in which you are interested. When you find a likely candidate, click on it!⁸
• When you arrive at the selected page, see if it satisfies your curiosity. If not, try the first step again. If it is the right subject but you are having trouble following the discussion because of unfamiliar terminology or concepts, click on the embedded hyperlinks in the text that will lead you to explanations of those terms or concepts,⁹ or invoke the MetaIndex¹⁰ to identify contextual links, on-line tutorials, quizzes, discussion groups or whatever other features have been incorporated into the HyperTextBook.¹¹ If your initial curiosity is satisfied but you would like to know how the current topic relates to others (in various different ways) or what sorts of mathematical tools might be useful in modelling the behaviour described ... or whatever, invoke the MetaIndex and have fun!
• Repeat the preceding step as many times as needed to achieve full satisfaction, or until you get bored and go away. (Curiosity is a powerful motivator, but it can get used up.) If you are unable to find what you need, use the handy "Need More Data!" form to submit a request for information to be incorporated.
• If, in your journeys through the HyperTextBook, you conceive of a connection that has not yet been recorded in the MetaIndex database, use the handy "New Link!" form to propose a new entry in the MetaIndex. You may also have personal knowledge that you feel should be incorporated as a new learning quantum, or which you think satisfies a "Need More Data!" request. (A list of the current backlog of "Need More Data!" requests can be viewed at any time by enthusiastic contributors.) Use the handy "My Two Cents!" form to submit your contribution directly to the "learning quantum input queue."¹²

What is needed most is a refined, easy-to-use mechanism for generating links and cross-references and storing them in a database which we call a MetaIndex which in turn includes methods for accessing this database from any page in the HyperTextBook and exploring links related by context, topical proximity or other criteria. This tool is essential for the resolution of the complex problem of generating the cross-linking scheme that makes a HyperTextBook out of a mere collection of "learning quanta."