Modular content creation
One is really taking advantage of the software by covering a topic, not with one big article, but with a network of smaller articles. For example, one could write a long 10,000 word essay on some famous person like Ben Franklin, which is no different from a traditional print article. Or, one could write an eight paragraph bio about Franklin, and then another article on each important aspect of his life: Ben Franklin/Childhood, Ben Franklin/Printer, Ben Franklin/Colonial representative in England, Ben Franklin/Paris years, Ben Franklin/Electricity, Ben Franklin/Bastard children. As these little articles are completed, they form a complete article, which is easier to reference because it is broken down into meaningful categories. It is also easier to update since the parts are separate from the whole, so work can be done on one section without effecting the others. And, it should be easier to create, because the work can be done piecemeal. You can start with a brief overview, then outline the important subtopics, then come back to these as time permits, or let others follow up your lead. We see this happening on the World War II page, where the major battles and figures of the era were listed.
This process of creating articles piecemeal is called Modular content creation.
This relates to the breadth vs. depth debate, because it shows that breadth, if seeded properly, will grow its own depth.
The grand and glorious modular content paradigm allows for and encourages a great deal of redundancy. At first this might seem like a bad thing, like we are squandering electrons or something. Then we remember that, unlike more traditional mediums, we are not confined by a limited number of pages, or even minutes, and so can be as redundant as we please. Why would redundancy be pleasing? Because the same fact or set of facts may have relevance in multiple contexts. For example, Ben Franklin's activities in Paris should be discussed in a treatment of his role in the establishment of the United States, but should also, one suspects, be considered in a treatment of his philandering. His work on electricity will have an important place in any treatment of his life, but also in articles about 18th century science, or late-colonial life. In traditional mediums, one is forced to organize information according to one plan: one can write a chronological history of Franklin from birth to death, or one can treat different aspects of his life separately, but not both. Here, we can do both, and anything else we think of. If every module is self-contained, then readers can obtain information within only the contexts in which they are interested. And the modules can be linked to and referenced only as needed, such that someone doing an article on the great philanderers could simply link to Ben Franklin/Bastard children and ignore all that business about science and politics.
An example of this is in American football. Each article is very short for the most part, consisting in many cases of only a definition of a term. If one already understands a term, he may read ahead in the body of one of the larger articles. If not, he can skip over to the term and return, thus getting an ever-deeper, ever-broader understanding of the subject. If one were to try to make a coherent whole of all that data in a single article and make it flow, it would be a daunting task. Done in bits, however, it was cake. AS
I agree that this is a nice advantage of hypertext, and we should use it when appropriate, but I'd caution about taking small chunking too far. As I said on Wikipedia commentary/Breadth and depth (in response to what someone else said), there are aspects of fields that consist of what might be called "narratives" that are rather difficult to break down into chunks. I would add here that it's sometimes (not always, but sometimes) very important, for understanding, that these chunks be kept together on the same page--a lot of people won't follow hyperlinks for fear of losing the thread of the "main" article or for failing to see the relevance of the linked article to the "main" article.
I think the football articles are a fine example of how "modular content creation" can work well. Another good example would be articles about basic technical terms in computer science and in biology. I also totally agree about the importance of relatively small chunks in order to be able to organize those chunks into larger articles organized in different ways; the example of Ben Franklin's discovery of electricity is great. It has a place in an article--no, a set of readings--about the American Revolution, but it also has a place in a set of readings about the history of scientific discovery. Actually, though, I'd rather see it labelled Ben Franklin's discovery of electricity rather than Ben Franklin/Discovery of electricity. Generally, subpages are messy, and I've decided I don't much like 'em except for very limited uses (e.g., "talk" pages). But, by the way, Lee, I think your poker pages do show one way to use them that makes sense. --LMS