In the May 2002 issue, JoAnn Hackos discussed why accidental reuse doesn’t work. Instead, information developers and instructional designers need to have a plan in place to develop content for reuse by design. The following are responses to the article.
I liked your article on accidental reuse. I think you are 100% correct. Accidental reuse does not work. I shall be very interested to see if anyone attempts to mount a counter argument in favor of accidental reuse.
I have always had difficulty with the way people talk about reuse, and I think it comes down to this: I see reuse as a means, not an end. The end is the elimination, or at least control, of redundancy. It is the existence of uncontrolled redundant information that makes it difficult to adapt content to changes in subject matter or requirements. It also inhibits the automated generation of information products, which limits the range of personalization, as well as restricting the development of certain useful kinds of navigation and linking schemes.
Redundancy can only be eliminated or controlled by conformance to a model and to a plan. The accidental elimination of redundancy is of no value, because you cannot rely on it.
Topic based reuse can be of value in controlling redundancy. However, redundant information can occur at any level of granularity. By organizing information into subject-oriented components (that is, organizing content on the content axis) and using markup to give access to the more granular elements of these components, we can eliminate or control much more redundancy.
We cannot eliminate all redundancy in content. However, we can eliminate redundancy completely for certain classes of information and control it reliably for other classes. This has the potential to deliver huge savings, and also to deliver a huge potential for quality improvement and the creation of multiple different information products.
There is, of course, a cost to moving beyond topic-based reuse onto the “content axis.” The model is more sophisticated and more particular. The modeling exercise is therefore more difficult, and more custom configuration of the CMS is also going to be required to support it. Finally, synthesis routines must be developed to create user-oriented topics from subject-oriented components.
On the other side, my experience to date indicates that such a system is actually easier to use for authors, once they adapt to it. Because more discipline is exercised by the information model, there is less overhead to the collaboration between authors, and authors actually have to know less about the system and how documents are built.
We could posit a reuse hierarchy:
…I just read the latest CIDM newsletter and could not agree more with your article on accidental reuse. It just does not happen. Reuse must be designed from the start and that is why successful reuse strategies require an organization that is working its way up the process maturity model. A level 1 or an early level 2 organization has little chance of successfully implementing an effective reusable content strategy.
…I think people will need a lot of guidance to be successful with a reusable object method of development. They can’t just write pure generic subject matter and hope it will fit together into a document, nor can they write a narrowly targeted document and hope the content will be reusable (as you so aptly pointed out in your accidental reuse article). Authors must design content to be reusable and the content must be designed to address an objective, either a learning objective or a performance objective.