December telephone seminar review: A brief, comprehensive indexing primer

by Alayna Schaffer and Donald Stoehr

In our telephone seminar of December 17, 2002, Seth Maislin of Focus Information Services succinctly summarized the three steps of all good indexing:

• Decide what's important
• Decide what to call it
• Decide where to put it

Seth expanded on his subject matter with a skillful overview of indexing. He presented nine short lessons, using material both comprehensible to rookies and interesting to more experienced indexers.

Seth's introduction emphasized four concepts that define the quality of an index:

• Is it technically accurate?
• Does it represent all information in the documentation?
• Does every representation in the index point to useful information in the documentation?
• Is the index intuitive, efficient, and approachable?

Many people do not realize that indexing can be considered as much art as science. For example, an indexer must first decide what to put into the index, then organize it creatively and appropriately. It is also somewhat surprising to note that while the indexer's knowledge of the subject matter and the audience directly affects the quality of the index, the author is often not the best candidate for an indexer.

An indexer has to keep in mind that readers often approach the documentation using different words than the ones used in the document for the same concepts and ideas. For this, Seth encourages indexers to "think outside the book."

Seth organized the material into specific topics, and used exercises for the participants to illustrate his points. The first exercise was "negative," asking the participants to index material that, we soon found out, did not warrant indexing. As his novel approach emphasized, trivial references only detract from a good index.

When deciding what references to index, Seth cautioned us not to include every occurrence of a word that seems important. Instead, include only occurrences of words that are surrounded by text amplifying the concept. Just because a word appears to be important, it may not justify inclusion in the index. At the same time, however, an index must be comprehensive and complete; if too many references are excluded as trivial, the resulting index will fall short of these goals.

The indexer must also consider labeling. This is a task for which the indexer's knowledge of the subject is critical. The vocabulary used to label concepts must be intuitive to the reader and at the same time be short, accurate, unambiguous, and functionally distinct from other neighboring labels. Seth emphasized that the indexer, in order to give the reader multiple avenues to content, must "exhaust the audience's vocabulary."

After labeling, placement of ideas must be determined. As Seth explained, ideas in the index can be placed almost anywhere. The effective indexer builds "access redundancy" by giving the reader several ways to locate desired subject matter. Strategies for access redundancy include:

• Indexing under every important word (the word must not be trivial and must have relevant context)
• Double posting (reversing hierarchical levels)
• Synonyms (placing ideas under synonyms)
• Categorizations (organizing by parent ideas)
• Cross-references (connecting related topics)

Once the content is defined, the index must be structured. A variety of structures can be used, including supersets and subsets, associated categories, and equal categories. As Seth indicated earlier, the labeling of each entry is important, but the context is critical. The key is making referenced information accessible. Seth's cardinal rule of accessibility is as follows:

• Ambiguity of content is bad
• Redundancy of navigation is good

Concept locators are important and valuable. Use locator formats carefully. When applicable, page ranges give the reader a strong sense of the depth of information associated with a reference. Avoid using special treatments, such as varying the formats of page numbers, annotating page numbers, and multi-part page numbers. These are typically more confusing than helpful to the reader.

The indexer may also need to consider that navigation might have to work both with and without page numbers, as in a single-sourcing environment. Seth stressed careful consideration of the number of navigational steps readers will tolerate in an online, pagination-free document.

Formatting rules are very flexible. As in any document, consider readability, easy recognition of cross-references, and overall consistency. In most cases, a no-frills approach is best.

According to Seth, automated indexing tools should be approached cautiously. Often index quality is sacrificed for improvements in the production process. For example, tools typically mark indexed references without regard to the surrounding text.

Overall, this telephone seminar proved an effective and cost-efficient means of furthering our index education. Seth's presentation style was well-suited to the format, and the handouts he provided were easy to follow and provided sufficient space for notes. He also used the question and answer sessions proficiently, responding to specific questions while making his answers widely applicable. Given this experience, we would certainly consider attending another STC-sponsored telephone seminar.