A Review of From Web to Workplace: Designing Open Hypermedia Systems
by Kaj Grønbæk & Randall H. Trigg
Table of Contents
MIT Press, 1999 (Digital Communication Series)
ISBN: 0262-07191-6. 386 pp. $47.95Review by Jim Porter
Case Western Reserve University
Prelude
Let me start by saying that I'm reviewing a book that wasn't written for me, and probably not for you either. (Who am I? I am an academic, a university professor specializing in rhetoric theory and professional writing. I do research in rhetoric theory, especially digital writing theory, and I teach graduate and undergraduate courses in rhetoric and technical communication.)
The authors of From Web to Workplace, Kaj Grønbæk and Randall Trigg, identify their intended audience as "researchers, professional system developers, teachers, and students who are interested in hypermedia design and the application of object-oriented design principles and frameworks" (xiv). The book is intended for hypermedia system designers, not for writing or even technical writing teachers.
Unless . . . unless . . . unless when we talk about hypermedia system design we are actually talking about "writing." Is web-based system design itself a form of writing, or will it become one? Where does "system design" leave off and "writing" take up? Much of what Grønbæk & Trigg talk about in their book--for example, system architecture--is clearly not the province of writing teachers and rhetoric theorists. Still, From Web to Workplace raises interesting questions about disciplinary territory in the digital age: what aspects of hypermedia system design are forms of writing? Which aspects should we become knowledgeable about in order to teach students hypermedia writing and, perhaps, to influence the design of hypermedia systems?
The common ground between system designers and writing teachers would seem to be the activity of professional writing at work--especially, how it happens in a computer-based work environment. We have to remember what system designers seldom acknowledge: in the workplace, computer systems are mainly writing and communication tools--and so understanding how they are used for writing/communication purposes is essential to the design process. Coming out of the CSCW (computer-supported cooperative work) movement , researchers like Grønbæk & Trigg think in terms of "work" rather than "writing" per se. But the research they are doing and the models they are developing impact professional writing in the workplace.
Open Hypermedia Systems and the Dexter Hypertext Reference Model
Key concepts in the text which are helpful to review include open hypermedia systems, the Dexter Hypertext Reference Model, basic principles of an open system, and the importance of collaborative work and linking.
Open Hypermedia Systems
Let's look at the vocabulary the authors use, starting with the book's subtitle. "Open hypermedia systems" refer to "advanced hypermedia structuring models and architectures that make hypermedia support available from the user's favorite computer applications" (3). Fundamentally, the authors are trying to develop hypermedia systems that are attuned to users' system needs; they want to make hypermedia systems more adaptable to what they call "the heterogeneity of information."
The authors advocate an open hypermedia approach--in fact, they offer a particular model, called Dexter-based hypermedia design--to improve hypermedia systems, particularly to help them deal with diverse information and systems in the workplace. They are interested in creating systems that support "professional writers" (my term, not theirs), especially their collaborative work involving document production, editing, and distribution. The problem, as the authors see it, is that in some ways the World Wide Web "constitutes a step backward" (316-317) as a hypermedia system. So Grønbæk & Trigg are developing a design model that can be joined with the WWW (the world's largest and most widely accessed hypermedia system) and with other sorts of hypermedia systems (e.g., corporate intranets) to make them more supportive of collaborative work.
The Dexter Hypertext Reference Model
The so-called Dexter Group was a group of hypermedia designers who first met at the Dexter Inn in Sunapee, New Hampshire in fall of 1988. Through several meetings in the late 1980s and early 1990s, this group began to develop a model and shared vocabulary for hypermedia design. From Web to Workplace arises from (but also attempts to revise and improve) the Dexter Hypertext Reference Model that the group developed. Grønbæk & Trigg have worked extensively with the Dexter Group, with researchers in the CSCW movement, and with researchers in participatory design. (They have, for instance, co-authored publications with Lucy Suchman, Susanne Bødker, and Morten Kyng.)
What is the significance of the Dexter Hypertext Reference Model? It attempts to make hypermedia systems more adaptable to users' contexts. For instance, let's say I create a WWW document with an embedded link in it (for example: filename.html#link) that allows you to jump to a specific section of that document (the place where I have inserted the "link" anchor). Fine so far. But what if you want to create your own embedded link to a different section of the document? If I am the owner of that document, only I can add in the target tags that would allow you to link to your desired location.
In the design that Grønbæk & Trigg advocate, "links can be named and assigned types, which the user can inspect and manipulate from the user interface" (51). In other words, in such an "open" system, the user can exercise more control over the information. We are not talking here about the user changing the original information or taking over the document. Rather, the system design allows the user to create his or her own tailored version of the document.
Principles of an Open System
"Tailorability" is one of the three principles of an open hypermedia system, the other two being "system integratability" and "application integratability." The principles say that a user's system should be able to integrate other applications and data formats and also be integrated into them as well. The authors cite Emacs (a text editor) as an example of an application that fits in well with other applications, while Netscape Navigator is criticized for not supporting "integration of its browser into other applications and environments" (10).
Collaborative Work
The authors complain that existing hypermedia systems "are generally still oriented to single users" (26). We don't yet have systems that are designed with cooperative work in mind. In Chapter 16, "Collaborative Hypermedia," the authors describe several existing programs that do provide support for multiple users (e.g., Intermedia)--but what they point out is that the WWW "represents a move away from hypermedia structures that can be dynamically edited and annotated" (224). The challenge for hypermedia design, one that they take up, is the design of collaborative hypermedia. Here is a place for potentially productive exchange between system designers and writing teachers. In this chapter the authors build a model based on lists of collaborative tasks; however, their model does not show much awareness of how collaborative writing happens.
Linking
Chapter 6 might be important for digital rhetoric theorists. In this chapter Grønbæk & Trigg consider and critique the metaphor of the "link," discuss its history (in hypermedia systems such as NoteCards and HyperCard), and point to its conceptual limitations. In digital rhetoric theory, people have applauded the link as a breakthrough, as a way to move beyond the limitations of the print page (e.g., Bolter). However, Grønbæk & Trigg point to the limitations of the link metaphor and the important variations between types of links. Conceptually, they are trying to break out of "the link-centric view of the world" that characterizes most hypermedia systems (61)--a view which leads toward a kind of monolithic, one-size-fits-all approach to hypermedia design. In this approach, the design forces the user to adapt to its parameters, rather than the design being flexible enough to accommodate the diversity of the user's system and materials.
System-Centered vs. User-Centered Approaches to System Design
Not many folks in my field (rhetoric, digital communication, technical writing) have examined the relationship between system design and writing. Two notable exceptions are Robert Johnson in his book User-Centered Technology (1998) and Johndan Johnson-Eilola in Nostalgic Angels (1997). If it weren't so disgustingly self serving, I would mention my co-authored book with Patricia Sullivan, Opening Spaces: Writing Technologies and Critical Research Practices (1997).
These works make connections between computer-based writing (or computer-based writing research or computer documentation) and the design of computer interfaces. Bob Johnson's book is particularly relevant to this review, as it provides a counterpoint to Grønbæk & Trigg. Johnson articulates the difference between "system-based design" and "user-centered design" (a more rhetorically alert approach, which a growing body of technical writing researchers, like Johnson, favor).
What is this distinction and why is it important? According to Johnson, system-centered design focuses on "the system's features removed from any context of use" (Johnson 122). In contrast, user-centered design begins with and focuses on "the localized situation within which the user resides" (Johnson 128). A user-centered approach to design does not assume that all users fit the same generic mold. User-centered design views audience diversity as central and significant (not as accidental or as derivative or as too complicated to address). It also views analysis of user context as the first, or at least concurrent, step in the design process. This difference between a user-centered (rhetorical) approach to design and a predominantly system-centered approach such as Grønbæk & Trigg's is significant in terms of our ability to design systems to meet users' (or writers') needs.
Overall, From Web to Workplace is heavily system-centered, though with some important concessions to a participatory or user-centered process model. That is, the authors focus on the design and structure of hypermedia systems MOSTLY apart from the context of use of particular (and different) users. To put this criticism a different way: their view is not a rhetorical one. They don't think very much about audience needs, apart from the computer system that the user has. They do refer frequently to "users"--but their view of users is generic; they don't distinguish between different types of users, different frameworks of use, different contexts.
In several places in the book, the authors do explicitly espouse a user-centered approach to design. The authors do cite Lucy Suchman (in fact, Trigg has collaborated with her on research), and Pelle Ehn. They do include a discussion of participatory design (277) . However, they don't cite Suchman's significant user-centered book, Plans and Situated Actions (1987). Most of their references seem to be courtesy citations. User-centered and participatory design principles do not substantively guide the framework or discussion. They don't consider variations among user types: differences in level of technical knowledge, differences in learning styles, differences in disciplinary orientation, and differences of use (e.g., workplace application), much less differences in age, race, class, gender, ethnicity, or culture. The key difference the authors focus on is the variation among platforms, operating systems, types of files and information, etc.--in other words, system differences.
One important exception is Grønbæk & Trigg's discussion of the Great Belt, Inc. project (in Chapters 2, 19, and 20), which involved their redesign of a hypermedia systemto support engineers' work.
I especially liked Chapter 2, a short (12 page) chapter which describes in some detail a case involving engineering work at a bridge construction site in Denmark. The case reveals in concrete terms how engineers use hypermedia systems to do their work--and how hypermedia systems do or do not support that work. What the case shows is the issue of working with multiple forms of data: different types of documents, different applications, different operating systems. Though the authors' focus in the book is not writing, the case description in Chapter 2 is interesting from a technical writing point of view, showing the complexity of the interrelationship between technology systems and technical writing.
The authors return to their case discussion again in Chapter 19, to consider the problems of document flow and document management at Great Belt. The goal of the authors was to help Great Belt develop a system to manage and track document flow, to coordinate complicated projects, and to annotate documents cooperatively. (As in the example mentioned above, the engineers at Great Belt wanted to be able to link to and from documents for which they did not have write access. The authors developed a link-based annotation system that allowed the engineers to do this.) In this part of the book, we clearly see the connection between what writers need (engineers working with documents) and what the hypermedia system can or can't do for them. In this discussion we see the open hypermedia design PROCESS--that is, we see how the design of the system arises from and addresses worker needs. In this chapter, Grønbæk & Trigg articulate a central principle of the design process: "Experimentation is a means of analyzing actual practice that supersedes formal explanations or espoused theories" (291). Here they acknowledge that the design process has to consider "actual practice"--but in other sections of the book, and throughout the book at large, the emphasis is system-centered.
Conclusion
In general, Grønbæk & Trigg are to be commended for attempting to make design innately more flexible for users--to give users more choices, more options, and more control over the flow and labeling of information. I agree with their critique of "individualistic" hypermedia systems, and I support their effort to imagine a form of hypermedia design suitable for collaborative work. Their approach to open system design is promising. I give credit to the authors for trying to think about user diversity, within their system-centered culture. In fact, compared to some computer science engineers I have worked with, they are keenly sensitive to user diversity. Grønbæk & Trigg are coming out of a system-centered disciplinary culture, but they are also influenced by developments in participatory design, they are working within the CSCW movement, and they are attempting to acknowledge real users' needs as important to system design. They are not as attuned to practice or to ethnographic methods as somebody like Lucy Suchman, say, but they are attempting to build a model which integrates user "experimentation" as key to the design process. We see this most clearly in Chapters 2, 19, and 20.
The problem with works like this one is that the authors are still mainly committed to a system-centered approach which leaves users a distant second. As writing teachers and rhetoric theorists we have to provide a countervailing force. We have to insist that what REALLY matters is people (audience), and we have to be advocates of user diversity and the actual lived reality of people working with and accessing hypermedia systems. These people are also professional writers and communicators, and so we have some stake in the design of those systems that will support their collaborative writing.
From Web to Workplace shows that there is still a mighty divide between what they do (system designers) and what we do (writing teachers, rhetoric theorists). We have to be willing to enter into a dialogue with system designers like Grønbæk & Trigg--to learn from them and also to teach them.
| Kaj Grønbæk is Associate Professor of Computer Science at Aarhus University (Denmark). |  |  | Randall Trigg is a researcher in the Systems and Practices Lab at Xerox Palo Alto Research Center. |
Their work, individually and collaboratively, focuses on the design of hypermedia systems to support cooperative work--in other words, CSCW, computer-supported cooperative work.
References
Bolter, Jay David. Writing Space: The Computer, Hypertext, and the History of Writing. Hillsdale, NJ: Erlbaum, 1991.
Ehn, Pelle. Work-oriented design of computer artifacts. Stockholm: Arbetslivscentrum, 1988.
Johnson, Robert R. User-Centered Technology: A Rhetorical Theory for Computers and Other Mundane Artifacts. Albany: SUNY Press, 1998.
Johnson-Eilola, Johndan. Nostalgic Angels: Rearticulating hypertext writing. Greenwich, CT: Ablex, 1997.
Suchman, Lucy. Plans and Situated Actions: The Problem of Human-Machine Communication. Cambridge: Cambridge UP, 1987.
Sullivan, Patricia, and James E. Porter. Opening Spaces: Writing Technologies and Critical Research Practices. Greenwich, CT: Ablex, 1997.