Designing records and document retrieval systems

ARMA Records Management Quarterly, Apr 1995 by Connelly, Jim

Several months ago, I was in the middle of a records management presentation for the senior management team of a large corporation. I had just launched into a rather technical explanation of document management and imaging technology. Suddenly my train of thought was shattered.

"Jim, would you stop calling it a records storage system?" interjected one of the Vice Presidents. "It should be called a records retrieval system."

"I'm sorry, but could you explain what you mean by that?" said I, stalling for time. No manager can resist that sort of question and I needed a moment to collect my thoughts. But to my amazement and gratitude, I was then given a very detailed explanation of what management really wanted.

Essentially, the management of this corporation believed that "Anyone can build a storage system, but few people can design a retrieval system"

For their complex subject files, they needed a simple system that they could use to retrieve specific documents. It had to key on the way they organized information in their heads and they way that they worked in their offices. The system's structure had to mirror their thought processes. It could not be cobbled together from bits and pieces of so-called "standard" systems. It could not be based simply on existing lists and patterns. They wanted their retrieval system to belong to the user.

Now remember that these people had no training in major types of classification systems. They didn't know when to use block-numeric systems, when to use encyclopedic, when to use duplex-numeric or middle digit for that matter. As records management professionals, we are aware of the advantages and disadvantages of some 10 to 15 types of systems. Often we select or consider selecting a system type from an analysis of records volume and complexity. Only then do we give thought to the actual design. But here was someone saying that the retrieval mechanism was critical and needed more of the user's input.

I could not agree more. Retrieval mechanisms are the key. How have records and document management professionals missed this concept? Why do we cling to ready made designs (off-the-rack solutions rather than custom-tailored)? Before we explore how retrieval mechanisms work, allow me to explain why I think we have missed some of the key ingredients for successful designs.

A SHORT HISTORY LESSON -- THE DESIGN TRADITION

Writing is a late invention of mankind. Pictographs were much more common than writing in our early history. It was only about the year 2000 B.C. that alphabets arose -- letters that depicted the actual sounds of words. Our Roman or latinized alphabet in part came from these early alphabets including the Phoenician. (Undoubtedly spelling became a problem shortly thereafter.) These alphabets soon became standardized. This standardization or uniformity of alphabets began to make recordkeeping in early civilizations less of an individualized storage method and more of a universal retrieval system.

In Egyptian times, records were kept on wax tablets, papyrus or stone. The record keepers used a system of hieroglyphs (a combination of pictures and symbols) for words. The Mayans also used glyphs for ideas, sounds and numbers. In both these cultures, only the priests and scribes were taught the system and as a result, these professions obtained a certain amount of job security.

The Inca civilization used systems of knotted, colored strings or "quipus" to keep their records. Each record keeper devised his own code which detailed what each knot. group of knots or colored string meant. Without the code, the records were worthless.

Many modern age "support staff" develop their own personalized (and indecipherable) filing systems. Even some records coordinators manage to design systems that are puzzling to the average worker. This is not an enviable achievement!

Many recordkeeping systems have been developed over the past 5,000 years. But it seems that these recordkeeping systems have not evolved very much. Why are we stuck on using personal systems? The answer to both this question and the question of how these systems can evolve lies in our brains and how they work.

WHY WE FORGET

In the 1970s, Endel Tulving and Joseph Psotka, two researchers from the University of Toronto, began to study memory from a unique point of view. They studied why we forget! Research has shown that as we reach the age of 30-40, our ability to recall information slows. These Canadian researchers began from the premise that memory depends upon two things:

* a memory trace that contains information about an event, person or object, and

* a retrieval cue, i.e., something in the immediate situation that reminds us of the past.

During their study, they discovered that over time either the memory trace deteriorated or there was no adequate retrieval cue to activate the memory. Their experiment was designed in part to discover whether one or both of these concepts was the crucial factor in memory loss.

From a records management perspective, their experiment is fascinating. They asked participants to review and essentially memorize lists of 24 words as illustrated by Table 1.