EPPO Topics Conform to a Type

One of the more unexpected characteristics of Every Page is Page One topics, is that they tend to conform to a type. We tend to think of topic typing as something specific to structured writing, something that is not natural to how we write, but something that is imposed on content for the sake of making it available for reuse, of for some other purpose.

But if we look at the millions of Every Page is Page One topics that exist everywhere across the web, we find that an very large number of them actually show strong characteristics of topic typing. Conforming to a type is actually something that comes naturally to an Every Page is Page One topic.

Most of the examples that we have examined in this series conform to a type to one extent or another. The most obvious is probably the recipe. It is not a surprise that recipes conform to a type. But what is worth noting is that the recipe type was not created by an OASIS working group. It has arisen naturally out of the experience of millions of cooks writing millions of recipes over centuries of civilization.

The used car review is another good example. A used car review may look like just a sequence of paragraphs. But if we look more carefully we will see that most used car reviews cover the same basic things: overview, equipment, notable features, interior and comfort, safety, economy, reliability, and price history. The order may differ, but pretty much every used car review will include the same basic information.

Once again, this type is not the product of a standards committee. Rather, it comes from the needs of used car shoppers. This is actually a consequence of another EPPO topic characteristic: having a specific limited purpose. Because a used car review serves a specific limited purpose for a used car shopper, it naturally covers a the standard set of information a used car shopper needs, and thus conforms naturally to a consistent type.

Next up is the Wikipedia article on Ottawa. One look at this topic will show that Wikipedia’s articles on cities conform to a complex and well defined type.  A sidebar covers issues such as geography, government, population, and time zone. The table of contents lays out the major divisions of the topics: History, Geography, Education, Economy, Culture, etc. Check out the entries on other cities and you will find a very similar structure.

Many other topic in Wikipedia show a similarly well defined topic types: vehicles, languages, flora, fauna, novels, and on and on. Simply browsing Wikipedia is an effective short course on topic typing. Once again, however, no standards committee established these topic types. Rather, they are the result of millions of readers and contributors building up different topics, filling in the gaps in the information, refactoring and refining their structure. They are crowd-sourced topic types, and thus topic types that do a great job of meeting reader’s needs.

That leaves us with two less obviously structured examples, Tom Johnson’s Faceted Classification, Faceted Search  and Bob DuCharme’s Push Pull Next. These are what, in the structured writing business, are called “concept” topics, which basically means a topic with no structure that you can usefully model in XML. But even here, there are some distinct patterns.

Each article starts by introducing a single concept (faceted classification, the models of XML processing), both then offer a number of examples that illustrate what they are talking about, and by restating the original concept and drawing some practical conclusions. Many “concept” topics follow this pattern, including this one.

Okay, its a pretty basic type, and one that most of us were taught to write in grade-school. But that is the point. When it comes to topics, adherence to a well established type is the norm, not the exception.

So, what is the practical conclusion that the type of this topic demands? Just this: there is nothing strange or difficult about writing topics to a defined type. Once you train yourself to write topics (as opposed to writing books in pieces), topic types will come naturally.In fact, you might consider this the proof that you really are writing topics, when they begin to naturally conform to their type.

The next step is then to capture these topic types in a structured way so that you begin to apply useful processing to them.