Introduction: Hypertext, Part 2

This is the second part of a two-part comparative review of books on hypertext and hypermedia. The first part (see CR, Review 9310-0766) discussed in detail five volumes that provide a general overview of hypermedia: Emily Berk and Joseph Devlin, Hypertext/hypermedia handbook; Jakob Nielsen, Hypertext and hypermedia; Roy Rada, Hypertext: from text to Expertext; Philip Seyer, Understanding hypertext: concepts and applications; and Nigel Woodhead, Hypertext & hypermedia: theory and applications. This part reviews books in three more focused categories: books that discuss a specific application area; those that are of special assistance to people who wish to implement hypermedia systems; and those that are collections of works from several authors rather than monographs.

Books with a clear focus on an application area

Six volumes focus on the implications of hypermedia for specific application areas. The two Barrett titles deal specifically with technical writing for the computer industry, and will be of interest to people who wish to use hypermedia for preparing technical documentation. Bolter, Landow, and Delany and Landow concentrate on the implications of hypermedia for humanists. All three of these volumes deal with how hypermedia changes the way people read and write, and point out fascinating parallels between hypermedia implementations and recent developments in literary theory. The volume edited by Delany and Landow also presents descriptions of a number of courseware applications, using hypermedia as a means of presenting material to students in the humanities. Jonassen and Mandl focus specifically on the role of hypermedia in teaching.

Barrett: Text, context, and hypertext; and The society of text

These volumes are the edited and expanded proceedings of two MIT annual conferences on Writing for the Computer Industry. They focus on the work of technical writers in preparing such things as computer documentation and training material. Increasingly, such material is being delivered in the form of hyperdocuments, and these conferences are thus beginning to explore the issues surrounding the application of hypermedia to computer documentation.

The 20 contributions to the earlier volume are grouped in three sections, dealing with the technology of hypertext, management of electronic publishing, and the design of online information. The technology papers discuss such subjects as the role of artificial intelligence and natural language processing in electronic documents; the functionality of document databases; and the use of object-oriented scripting languages. The management section addresses the kind of people issues that too frequently keep new technologies from realizing their promise. It offers papers on coordinating document development with product development; managing organizational and turf issues; quality assurance in electronic documents; in-house training; and guidance on how to start an electronic document project. The papers in Part3, “Designing Online Information,” discuss style issues for hypermedia documentation; the appropriate application of different genres (from low complexity, adjunct material like system messages to self-contained, highly complex systems like computer-based training); constraints imposed by the physical computer screen and how to accommodate them; and techniques for writing for a multinational audience.

The 23 papers in the more recent volume cover the same three basic areas and add a fourth: how to design systems for online users. The four papers in this new section describe how online information resources differ from human ones; how to assess user wants and needs; and the role of artificial intelligence in making systems more usable. The technology section in The society of text is dominated by descriptions of particular systems, with a focus on different hypermedia architectures: Athena Muse, Intermedia, Hyperties, the authoring tools in Symbolics’s Concordia, NoteCards (as a prototype of a collaborative system), and Apollo’s Domain/Delphi. The section on managing technical writers emphasizes the skills that the new paradigm requires of individual technical writers, including consulting skills and the ability to work collaboratively. As in Text, context, and hypertext, the last section focuses on style and usage issues, but has a special emphasis on the impact of an online environment for collaborative work.

Both volumes have integrated indices. Neither one has an integrated bibliography.

Bolter

Bolter examines the characteristics of the various spaces within which writing may take place, and discusses the implications of the computer as such a space, in the context of recent literary theories.

An introductory chapter discusses the shift from printed paper to electronic media for literature, and anticipates the major argument of the work, which is that the shift in media will lead to new and different capabilities and experiences. A two-page concluding chapter does not summarize the conclusions, but rather points out the inadequacies of the printed medium for the discussion and invites readers to explore the electronic version of the work, available as a Macintosh diskette. The intervening 12 chapters are grouped into three parts, discussing the impact of the computer on the visual and conceptual spaces in which writing takes place and then extending these insights to inquire how the new modes of reading and writing will change the way we think.

Part 1, “The Visual Writing Space,” explores the technology of writing, both conventional and electronic. Chapter2 is the obligatory overview of the concepts and history of hypermedia. Chapter 3 invites the reader to see writing as a technology, and traces the progress in that technology through history up to the electronic age. Chapter 4 gleans from this history a number of elements of writing, including pictorial, phonetic, and iconic approaches and the various techniques people have used to lessen the finality of a written document. Chapter 5 focuses on the visual impact of writing, beginning with typography and embracing electronic screen design and the use of diagrams and graphs.

Part 2, “The Conceptual Writing Space,” moves away from a strictly visual emphasis to deal with how writers manipulate concepts, developing the impact of electronic writing on four literary metaphors: book and library (chapter6), dialogue (chapter 7), interactive fiction (chapter 8), and critical theory (chapter 9).

The three chapters in Part 3, “The Mind as a Writing Space,” argue that the changes that electronic documents bring about in how we read and write will lead to changes in how we conceive of ourselves as readers and writers, and thus electronic writing provides a new metaphor for thought and culture.

Bolter’s essay is not an engineering book. It does not tell readers how to build a hypermedia system. However, engineers should read it for the insight it gives on the potential impact of electronic publishing technologies on human communication and thought in general.

The volume includes a general index and a bibliography through 1990.

Landow

Landow’s monograph overlaps heavily with Bolter’s in its concern for the literary and humanistic impacts of hypermedia. As the subtitle indicates, Landow emphasizes critical theory, which Bolter treats mostly in a single chapter in the context of a broader theory of writing.

Landow’s dominant thesis, developed in four central chapters, is that hypermedia reconfigures four aspects of literature: the text itself, the role and activity of the author, the nature of narrative, and what happens in literary education. The text itself, with the addition of graphical icons, immediate nonlinear associations via links, closer reader control over the sequence of presentation, and the possibility of nontextual material, is radically changed. The distinction between author and reader is greatly diminished, and new mechanisms are available for collaborative writing. The nonlinearity of hypermedia challenges the most basic elements of narrative plot, and requires the development of a new set of structures. In the instructional context, students gain power and authority with respect to the instructor as the technology permits them to access and relate sources more readily; students can work more at their own pace, so the class schedule becomes less important; and the notion of the literary canon becomes much vaguer as documents are accessed electronically rather than constrained between bound covers.

An overview chapter introduces hypermedia and relates it to concepts in critical theory. Chapter 6 discusses political issues raised by electronic text, including questions of access and the reduced authority of the author. The conclusion, as in Bolton’s work, is a brief, one-page restatement of the urgency of the book’s themes rather than a synopsis of those themes.

The book includes a bibliography through 1991 and an index.

Delany and Landow

This volume of 17 essays was compiled from contributions solicited via email from people in the humanities. The contributions are grouped into three parts: “Introduction,” “Theory,” and “Applications.”

Part 1 is a lengthy (50-page) introductory essay by the editors that surveys and synthesizes the material in the later contributions. This essay is in many ways a sketch or prototype of Landow’s fuller discussion in his monograph, discussed above.

The eight essays in Part 2, on theory, sound many of the same themes as the monographs by Landow and Bolter discussed above, and Bolter’s contribution here stands in the same relation to his monograph as Landow’s introduction does to his. The concepts covered range from fairly abstruse discussions of literary theory (deconstructionism, decentering, and intertextuality) and a psychoanalytic model of textual digression, to concrete rhetorical rules for authors and guidelines for users of hyperdocuments.

Part 3, “Applications,” gives detailed descriptions of seven major applications of hypermedia to literary studies, including biblical studies (the CD-Word project of Dallas Seminary); classics (Harvard’s Perseus project); a corpus of Chinese documents and supporting literature (in the Intermedia framework); performances of Shakespeare (via videodisk); the European emblem literature; Fielding’s Joseph Andrews; and a growing collection of works by Ruben Dario. Each of these essays reflects explicitly on principles and lessons learned from the effort. An eighth essay develops a conceptual framework for implementing educational hypermedia programs.

The collection is well integrated, beginning with the introductory synthesis and extending through the global bibliography (through 1990) and overall index.

Jonassen and Mandl

These revised papers from a 1989 NATO Advanced Research Workshop focus on the design, implementation, and use of hypermedia for instructional purposes. The 24 contributed papers are grouped under six categories: hypermedia in learning; designing the information model; designing the user interface; hypermedia and instruction; the hypermedia design process; and conceptual foundations for designing hypermedia systems for learning. An introductory preface develops a model of the psychophysics of hypermedia as a tetrahedron whose four vertices are the user or learner; the domain or materials being studied; intellectual and/or learning activities; and the motive for hypermedia use.

The most distinctive feature of the book is its attempt to simulate hypermedia in print. After the revised papers were collected, the entire volume was sent to each participant for comment, with encouragement to indicate comments in the form of links among articles. These links are typed, with functions such as “example,” “contrast,” and “results.” The final edition prints these links in a wide margin alongside the text, so that readers can quickly access related material. Each paper begins with keywords and an analytical outline, further facilitating rapid review of the contents.

As is often the case in proceedings, there is no common bibliography or index.

Help for builders of systems

Two groups of volumes will appeal to people who understand the concept and history of hypermedia and need to engineer systems. Five volumes (Murray, Nielsen, PDR, Seyer, and Woodhead) review several commercially available systems, often classifying them according to their intended use. They provide a good starting point for selecting a shell for a new application. Part 1 of this review discussed Nielsen, Seyer, and Woodhead. Five volumes (Murray, Martin, Horn, and the two Barrett books) give specific direction on the rhetoric of hypermedia, to guide writers in organizing and dividing their material to help readers avoid confusion and quickly find what they need.

Murray

Murray’s focus is on text-oriented hypermedia. He writes in a friendly and conversational style. While the title of the work focuses on building hyperdocuments beginning with documents in Ventura format, the insights are much more broadly applicable. The book leaves the reader with a view from the trenches from someone who has been there. While the volume is not as glossy as some others, it gives concrete, specific advice that cuts to the heart of many critical development challenges.

The material is organized into eight chapters (they are labeled “parts” in the book). The first two parts give a basic definition of hypermedia, the author’s personal experience in entering the field, a brief history of the field, and general architectural features of hypermedia systems. A sidebar in Part2 provides a helpful classification of various commercial systems into seven categories: authoring environments, trans-application linking services (enabling users to link different applications to one another), knowledge analysis/generation systems, hypertext publishing systems, global knowledgebase systems (at present represented only by Xanadu), expert systems with hypertext features, and text retrieval systems with hypertext features.

The next two parts equip the reader to decide whether to use hypermedia for a given application. Part 3 outlines the benefits hypermedia offers to authors and users, and Part 4 discusses different kinds of documents and the relative ease and benefits of converting them into hyperdocuments.

The next three parts provide technical direction in actually producing a hyperdocument. Part 5 discusses the selection of a particular hypermedia system, providing a list of criteria that a document designer should take into account. Part 6 walks the reader through the mechanics of converting text to hypermedia. While the examples are specific to the Ventura context, the issues raised apply in any development environment. Part 7 explains the role of a data model in integrating organizational information, and urges the reader to take a systematic rather than a spotty approach to implementing hypermedia.

Part 8, the conclusion, identifies areas in which hypermedia still needs to grow as a technology, and speculates about its future. An appendix gives contact information for 43 commercial vendors of systems mentioned in the book. The volume is thoroughly indexed.

Martin

Martin’s volume follows the style and approach of his other “how-to” books on computer systems: a broad technical overview that avoids research depth; an emphasis on business and management considerations; and numerous explicit principles called out clearly. The book’s 17 substantive chapters are organized into four parts.

The three chapters in Part 1 introduce the concept of hypermedia, discuss its benefits with respect to particular applications, and consider how it can be integrated with other computing techniques such as databases and expert systems.

Part 2 offers five chapters on the structural organization of hyperdocuments, while Part 3 provides six chapters of stylistic advice for authors. Part 2 focuses on what happens between nodes, while Part 3 considers the contents of an individual node. Both sections give special attention to graphic as well as textual material. The recommendations of these sections, while clearly stated and motivated, do not take into account the capabilities and limitations of specific commercial offerings, and thus do not help the user identify a specific development system.

The three chapters in Part 4 consider the organizational and managerial issues that arise with electronic documents. Because they are dynamic, they involve maintenance considerations that do not arise with paper documents. Because they are often produced by groups of authors, they require special organizational structures. The section gives special attention to the process of converting paper documents into electronic form.

The coverage of the book, in particular the emphasis on management issues, is similar to the two volumes edited by Barrett, but Martin’s book is more integrated and less technical, befitting the difference between a monograph and an edited proceedings. The limited references are segmented by chapter rather than integrated into a single bibliography, but the book is indexed.

PDR

This report is an in-depth review of 40 hypermedia systems for PCs, and will be an invaluable reference for users just beginning to select a system. The first 70 pages describe hypermedia, review its history, establish terminology, and identify evaluation criteria. The next 230 pages discuss the systems reviewed in six categories: document development products, help systems, pushbutton database/programming systems, expert systems, text retrieval systems, and personal information managers. For each product, the review gives an overview with pricing; the central design metaphor; a discussion of how the system describes and manipulates nodes, elements, documents, and anchors; support for five different kinds of links; usability features for end users; development features; operational features; expected new versions and features; a summary recommendation; pointers to other published reviews; and information on the system vendor.

Horn

This volume applies Horn’s information mapping method of decomposing and packaging material to the preparation of hypermedia. The book is highly visual, with pages that look more like a multiwindow workstation than a printed document, and with explicit links connecting discontinuous sections of the text. The nine chapters are organized into four parts.

The two chapters in Part 1 introduce the reader to the basic concepts of hypermedia, and identify a number of issues that must be resolved in constructing a hyperdocument, including the contents of nodes and links, the problem of cognitive overload, maintenance difficulties, navigation, the distinction between holistic and serialist readers, and the need for new skills and new rhetoric.

Part 2 offers three chapters on information mapping. This technique, which has been applied for 20 years in preparing paper documents, is a method for defining and organizing small chunks of information called information blocks. The method identifies different kinds of blocks. These blocks are grouped into information maps (for example, an illustration and its textual description). This volume disclaims any complete presentation of information mapping, but refers the reader to other publications and courses offered by the author’s company. Relationships among blocks and maps are defined by hypertrails (sets of links that organize and sequence information from a particular perspective, such as chronology, decision, structure, geography, or classification). The last chapter in Part 2 discusses how this organizational strategy addresses the challenges raised in Part1.

Part 3 offers case studies and examples around three kinds of literature: stable discourse (documentation and training), disputed discourse (analyzing arguments), and experimental discourse (scientific material). The single chapter in Part 4 looks ahead to the implications of integrating communications with hypermedia and the impact of virtual reality.

An appendix gives a two-page overview of each of several pioneers and exploratory systems: Vannevar Bush, Doug  Engelbart,  Ted Nelson, Andries Van Dam, the Zog at Carnegie Mellon University, the Massachusetts Institute of Technology’s Dataland, Peter Brown’s Guide system, John Sculley and Apple’s vision for a Knowledge Navigator, and Bill Atkinson’s HyperCard.

Multiauthor

Nine volumes (the two Barrett books; Berk and Devlin; Delany and Landow; the Hypertext ’87 [1], ’89, and ’91 proceedings; Streitz et al.’s Proceedings of the European Conference on Hypertext; and Jonassen and Mandl) are collections of papers by multiple authors. Four hypertext volumes are traditional conference proceedings. Barrett’s two volumes and the volume by Jonassen and Mandl also originated with conference papers, but have extensive editing and arrangement. The other two volumes, edited by Berk and Devlin (see part 1 of this review) and by Delany and Landow, are more deliberately engineered collections that seek to bring together contributions of many authors to tell a single consistent story. In general, all of these volumes offer a deeper technical treatment than do the monographs, and provide access to more up-to-date research.

The four hypertext proceedings volumes provide an interesting perspective on shifting emphases in the field over the last five years. To provide an overall summary, the 139 contributions in these volumes (including papers, panel summaries, and descriptions of video presentations) were classified into five categories: general overviews of the field (typically, keynote addresses by invited speakers); underlying technologies and models for data, structure, and navigation; descriptions of systems independent of a particular application; discussions of specific applications; and issues of situated use (such as authoring, document conversion, and reader support). The relative emphasis on these five categories progresses through the history of the hypertext conferences. The initial conference (Hypertext ’87 [1]) emphasized systems (such as KMS, Guide, and Symbolics’s Document Examiner) and complete applications (such as the electronic Oxford English Dictionary), with almost no attention to enabling technologies. This tendency is expected early in the life of a field. Researchers are exploring the space of possibilities through prototype systems. Two years later, at Hypertext ’89, people were focusing their attention more on the underlying technologies, such as navigational mechanisms and refined definitions of what constitutes a link. At the European conference in 1990, the emphasis on issues of situated use reflects a general European strength in matters of ergonomics and job design. The Hypertext ’91 conference differed from the three previous conferences in having no strongly dominant or neglected theme. The field has become mature enough that reasonable research is being done evenly across the spectrum of relevant topics.

Summary

The rich variety of books available on the subject of hypertext and hypermedia reflects the vigorous activity in the field, both in research and in the commercial exploitation of this powerful metaphor for human-machine interaction. Readers wishing a general introduction to the field will turn profitably to Berk and Devlin, Nielsen, Rada, Seyer, or Woodhead. Those who wish to focus on research issues will find Rada and the various conference proceedings most appropriate. People who wish to build systems have ample guidance in the works discussed in this part of the review.