If enough people take this book seriously, the field of technical communication will be turned on its head. James’s book is an outright challenge to documentation managers, printing managers, software suppliers, and technical writers to bring the technical documentation world into the Information Age. Word processing, desktop publishing, faster computers, etc. are not enough, he says--if new technology is not tied to new methodology, writers will only jam up their systems faster.

A document database is a collection of chapters, tables, and artwork, each in its own file. In James’s ideal system, files can be pulled together to create anything from a brochure and help screens to mechanical specifications and full-scale technical manuals. None of the formatting information is kept in the individual files. Instead, printing or display instructions are added at the last minute.

The advantages of the system, according to the author, are as follows:

• All of the information is available at any time. It can be accessed or printed both at the client’s site and in-house.

• Since there is only one copy of any graphic or piece of text, changes are much easier to control.

• It simplifies the interactions among writers, programmers, and clients, raising the quality of the material.

• The publication manager or client can print only those parts of the total database that he or she needs. As the author explains in Chapter 1, Methodology, a widget user’s guide can be customized as easily as the widget itself.

For example, in Chapter 2, Text Manipulation, the author evaluates several word processing systems. (He does not use brand names, which is appropriate because the range is broad. I suspect, however, that his readers will henceforth scan software ads more critically.) In Chapter 3, Graphics Manipulation, he explains the differences between vector and raster graphics (a valuable bit of information), as well as how graphics and font generators usually work. Chapter 4, On-Line Document Querying, covers types of online databases, request methods, and access methods. In Chapter 5, Indexing, he describes the virtues and failings of automatic and declarative indexes.

From Chapter 6 on, pointed comparisons are made between traditional and document-database methods. In Chapter 6, Distribution and Maintenance, the author compares the standard book-oriented distribution system to the new electronic distribution methods. The primary disadvantage of the first, he says, is how difficult reader feedback becomes: “Nearly every standard textbook on writing techniques stresses the importance of knowing the intended audience. [Therefore], it is perhaps surprising that so little effort is actually taken to allow the reader to communicate with the writer. . . . It is almost as if the technical writer . . . takes the manuscript and throws it over a wall to the readers” (pp. 116–117).

In Chapter 7, Quality Control, the author plugs the new documentation tools--automatic spelling checkers, readability tests, style checkers, etc.--and explains why controlling changes is easier in document database systems than in others. In Chapter 8, Systems, he presents hardware and communication configurations that one might use to implement a document database. Unfortunately, his description of a “non-integrated workflow” seems unlikely; he’s already out of date on electronic page makeup, and his ideal system, the “integrated workflow,” is written as if it may be only theoretical. But these are small quibbles.

Chapter 9, Productivity and Environment, is the glory of the book. Here, the author demonstrates just how interactive an electronic system can be with a “day in the writer’s life” example--it’s enough to make one drool. He also compares his system’s productivity and cost performances with those of traditional technical documentation systems: “The document database is slightly over three times more productive than the industry average” for the same amount of money (p. 159).

Nevertheless, the book is flawed by two prejudices. First, the author does not believe in aesthetics: “Electronic page makeup is designed for entertainment publications rather than technical publications. . . . With technical publications, however, the problem of keeping the information accurate is far more important than the prettiness of the page” (p. 152). When your hardware and software cannot generate attractive pages, then perhaps you are justified in concentrating on what it does do well. But all other things being equal, why throw out beauty? James is wrong to disregard the results of more than 20 centuries of research into making books easier to read.

Second, the author does not believe in editors: “automatic spelling and grammar checking greatly reduce the need for proofreaders. Because declarative structuring automatically enforces publications standards, copy editors and format editors are no longer required.” Proofreaders do not check grammar--editors do. In addition, no self-respecting human editor would have let him get away with “Changes are now taking place that obsolete [reviewer’s emphasis] the communications techniques of the past” without a big fight (p. 1). You can, on the other hand, always turn off an automatic style checker, which may have been what he did. Finally, a copy editor would have caught the nasty confusion on p. 160: “the proportion of technical communicators to programmers was better than 6 to 1, as opposed to 3 to 1 before the use of RUNOFF.” He meant 1 to 6 and 1 to 3. (Some day this reviewer would like to see a book by a writer on how to program--an appropriate response to all these books by programmers on how to write.)

Readers can safely skip Chapter 10, The Future. Events have already overtaken and disproved him (see [1]). What this chapter should have contained was help for the author’s new converts. Surely his document database system can be implemented on systems other than the Honeywell Control Program-6 Operating System. UNIX/XENIX development systems may have some of the necessary file- and text-handling facilities, for example, and the latest multi-user operating systems are supposed to open new worlds for microcomputer users. But without samples of code and detailed flow charts, programmers and computer-smart technical writers will have to reinvent the entire system for themselves. Or will we see Document databases--the sequel soon? I hope so. This book deserves to be followed up.