Book Review of

Mapping Hypertext

The Analysis, Organization, and Display of Knowledge for
the Next Generation of On-Line Text and Graphics


Robert E. Horn

Reviewed by Carl Binder

Reprinted by Permission from Performance and Instruction (a publication of the National Society for Performance and Instruction) , October 1991


Mapping Hypertext by Robert E. Horn is a tour de force in several respects. First, it is an amazing example of "graphic language," the use of complex graphics as an inherent part of the communication. In fact, this book is as close as a paper-based document can be to hyper-graphics. It simulates the rich graphical environment of current-day and future object-oriented computer workstations. Second, Horn's book is a desktop publishing miracle. Created entirely on the Macintosh computer with MacDraw II, it is a remarkable display of textual and graphic information, including hundreds of icons, complex drawings, tables, and structured text. The author deserves some kind of award for design and development of the document itself, independent of its content. Finally, as a contribution to literature, Mapping Hypertext is a unique and seminal work, covering the history and conceptual underpinnings of hypertext, suggesting applications and design principles capable of stimulating hypertext and hypermedia design for years to come, and providing hooks to the Information Mapping method, a systematic approach that promises to take much of the guess-work out of hypertext development.

Mapping Hypertext was a finalist for the 1991 Outstanding Instructional Communication Award, and for good reason. What fewer people may know is that Horn won the NSPI Outstanding Research Award in 1976 for the work which led to his trademarked Information Mapping method. This book continues in a tradition of turning basic research into useful performance technology, an endeavor that has kept Bob Horn busy for over 25 years. He was an early participant in the programmed instruction movement, the field which gave birth to NSPI. Due to proliferation of poorly designed, untested and ineffective "programmed learning" materials during the 1960's and 1970's, programmed instruction itself lapsed into some disrepute. When the idea of programmed instruction caught on with publishers and authors, many jumped at the chance to make a buck. In their haste, however, many developers ignored basic programmed instruction principles of systematic analysis, design, development, evaluation, and revision.

Current excitement about hypertext, hypermedia, and hyper-everything-else threatens to undermine the potential value of this new technology in much the same way: by encouraging "creative" development without attention to systematic principles or effectiveness. Horn's book and the Information Mapping method offer a clear path through the maze of potential hyper-junk.

Overview of the Book

The book contains nine chapters and an appendix sorted into four parts. Part 1: Hypertext - Hypermedia, New Opportunities, introduces basic concepts and reviews current issues in the field, providing plain English definitions, historical notes and lots of compelling graphic illustrations. Part 2: The Method of Information Mapping, summarizes basic principles of the Information Mapping method and shows how these principles offer solutions for key navigational and design problems in hypertext. Part 3: Some Applications of Structured Hypertext, illustrates how these principles might apply to hypertext documentation in three general areas: documentation and training, argumentation analysis, and experimental scientific discourse. Part 4: Mapping Future Infospace, summarizes the book's key arguments, projects trends and likely future developments, and provides an appendix covering most of the key historical figures and developments in the field.

A Few Observations

This is not really a "how to" book, although it provides a rich set of ideas, guidelines, principles and suggestions for how to design good hypertext. It is not a procedural text on the Information Mapping method itself, and this might frustrate some readers who want to know more. Horn's company, Information Mapping, Inc., provides in-depth seminars on how to "do" Information Mapping documents. Those who want to master the method should attend these seminars (just as those wishing to master criterion referenced instruction should attend Mager's workshops, for example). On the other hand, the range of issues and topics covered by this book is so comprehensive that any prospective or experienced hypertext developer should acquire it. Developers will profit from reading and re-reading it, using it as a guide, and keeping it nearby during crucial stages of design and development.

Because the book is a simulation of hypertext developed according to the Information Mapping method, it is full of "chunks" organized hierarchically into larger contexts. Many of these chunks are nuggets, intriguing or useful as the case may be, for even the casual browser. For example, in discussing the importance of identifying the types of information that appear in different domains of discourse (or subject matters), Horn points out that, "an analysis of domains usually finds that there are similar clusters of information block types across many documents of the same domain. This enables us to clarify the similarities of each of the information block types in the new domain and thus to write and to teach the writing of these documents much more precisely and easily" (p. 105). This observation is perhaps obvious, once understood; but like many important truths, it has powerful implications. In fact, much of the power of the Information Mapping method rests on an analysis and classification of types of information, an endeavor that supports what Horn calls "precision modularity." Because good hypertext is inherently modular, an approach that identifies basic units of information analysis provides power tools for developers. Precision modularity allows developers to create user-friendly hypertexts that are comprehensible and navigable because of their consistency, structure, and manageability. Most books and articles on hypertext and hypermedia seem extreme in one direction or the other. Either they contain craftsmanlike "guidelines and rules of thumb," or they are severely academic, discussing theories of cognitive structure, principles of user-interface design, linguistic analyses of hypertext, etc. This book, like the Information Mapping method itself, cuts a fine line between the two extremes. While it refers frequently to relevant research (e.g., on chunking and the limits of short-term memory), it also describes useful processes and guidelines for information management and the design and development of documentation. For example, a two-page text and graphic display describes how four principles of Information Mapping constrain the construction of information blocks-- the basic units of information in a document (p. 86-87). These pages provide practical, powerful guidelines for defining the building-blocks of a document. For those inclined toward metaphorical or futuristic thinking, or right-brained imagination, there are plenty of expansive graphic models and descriptions of future hypertext systems. Much of what Horn does is to provide a logical extension from current day hypertext capabilities and applications to future possibilities, based on a systematic analysis.

Among the most interesting chapters in the book is "Navigating Structured Hypertrails" which presents a classification of author-defined sequences through hypertext. Horn defines a hypertrail as "a set of links between chunks of information...that organize and sequence information about a particular function or characteristic of the subject matter." He describes "linearizing hypertrails" as "a process of selecting from the networks of links a series that makes sense as the primary hypertrail for the user." In other words, linearizing a hypertrail is a way of defining what happens when the reader selects the <NEXT> button. As Horn points out, a single hypertext document may contain multiple hypertrails, defined by different organizing principles. His classification of hypertrails includes those sequenced according to prerequisite information, classification schemes, chronology, geography, stages in a project, structural elements, steps or options in a decision process, definitions, and examples. One could probably think of others. The point is that by classifying the types of paths a reader might wish to take through a body of information, a hypertext designer can better address the informational and performance needs of one or more audience types.

Without a doubt, the Information Mapping method provides tools for resolving major hypertext reading and writing problems. In its systematic analysis of user performance requirements and types of information, the method takes the readership into account. By applying principles of chunking, relevance, labeling, and consistency, the method helps manage the limits of short term memory and comprehension. By incorporating information management principles such as hierarchical structure and advanced organization, the method enables authors to help readers navigate through what otherwise might be a maze of hyperspace. Although good writers and designers may already use many of the principles incorporated in the Information Mapping method, no other system (to my knowledge) organizes and systematizes such a complete set of principles in a way that writers and designers can apply them so consistently and comprehensively.


So what can we say, in conclusion, about this book? First, it is an imaginative and adventurous exploration of conceptually and technologically interesting territory. Second, it is an exemplar of graphic communication in conventional paper form. Third, although it might sometimes frustrate the reader by telling what needs to be done to make good hypertext but not always describing exactly how to do it, Mapping Hypertext is nonetheless an eminently practical book, full of good counsel for hands-on developers. Finally, although one might at points disagree with some of Horn's extrapolations or suggestions for the future, one must be impressed by the depth of his analysis. For anyone even slightly interested in hypertext or hypermedia, this is a must-have volume.

About the Reviewer

Carl Binder is President of Precision Teaching and Management Systems, Inc., and Co-Chair of the NSPI Emerging Technology Committee. Contact him at PT/MS, Inc., P.O. Box 169, Nonantum, MA 02195 (617)332-2656 or CompuServe 73240,1134.

Information Mapping is a registered service mark of Information Mapping, Inc. of Waltham, Massachusetts.

R. Horn Home