Wednesday, June 21, 2017

Notes on pattern-making: June 21, 2017

I've been continuing to read and learn what I can about pattern languages. My questions are: How do I identify a pattern? What is the difference between a taxonomy and a language? How many patterns are too many? What do I want to do with this thing?

Luck for me, some research on these very questions exists on the web for anyone willing to do the reading and to tolerate the jargon of other disciplines. What follows is a disorderly brain dump of notes that I have taken so far.


"While individual patterns are useful on their own, they are more powerful when three are used in combination."

To start the process, an initial design problem is chosen and the designer applies one or more related patterns to solve the problem. The process continues as the designer considers additional problems that arise from the requirements and constraints of the project. Additional patterns are applied until the designer is satisfied that all design problems have been adequately addressed through the application of patterns. Alexander's descriptive framework for outlining a pattern included a pattern name, a statement of the problem underlying the pattern, the context in which the problem occurs, one or more solutions to the problem, and any related information that might aid others in connecting the pattern with other related patterns.

as new patterns were generated, previously written patterns were revised in order to clarify any new relationships that were discovered to exist between the patterns.

patterns are considered mature as they explain their topic and suggest ways of encouraging the quality without a name to a greater degree.

The mere fact that Alexander used the descriptive phrase “the quality without a name” to name the quality he wished to describe is somewhat ironic, and this may confuse some readers. After all, if it does not have a name, can it really exist? However, readers should understand that Alexander used that phrase in part to emphasize the precise nature of the phenomenon and remind the reader that when they experience this quality they may have difficulty communicating the entire experience through individual words. Nevertheless, even though it may be difficult to describe this quality and to understand it intellectually, it should be remembered that Alexander claimed this quality is real and that it can be readily identified by asking the simple question of whether or not the artifact, event, or experience feels good.


The emphasis that practitioners have placed on Alexander's "quality without a name" has been a surprise. Alexander used this quality as a telos or target by which to judge the fitness of his entire project. And others have done the same after taking time to describe what this quality may be in their discipline. Their attention to this detail makes me ask whether I also shouldn't describe a quality for my own project.

My instincts gestures toward a kind of religious depth. Perhaps this is akin to what Schleiermacher meant by a "feeling of absolute dependence." This is the kind of thing William James is getting at in his famous lectures, or that his peers, the Transcendentalists, meant by the Oversoul. This is the sort of company that yanks me to a stop. Awe toward a transcendental mood is a false positive. So, I think that I will go a different route and call it "worship." The "quality without a name" my pattern language is aiming at is demarcated at this point by the word worship.

Moving on, I wonder if orthodoxy is a descriptor that means that certain patterns are present (or absent) in a matrix? And I wonder if the Trinity works as a kind of meta pattern or organizing principles that acts as a heuristic to the whole. In other words, if your pattern organizes itself according to a trinitarian heuristic, then you are working within the bounds of orthodoxy.

Asking after the transition from bare taxonomy to pattern language, the turned corner of difference seems to be organization along x, y axes. The x are related cards. The y is meta-relationships, both up into the abstract and down into the particular. I picture at least two levels and perhaps three. And so, I have to ask what my abstract index layer will look like. How complex will it be? How complex must it be to cover the field? And what qualifies a topic to exist at this level? Is this a simple matter of writing Father, Son, and Spirit on three cards and going for it? Or, will these meta principles be much more abstract, such as "being poured out"?

Finally, the question of what to do with this tool is slowly answering itself as work progresses. Here are three applications that I've been thinking of:

  • opening doors. A theological pattern language will help an individual or group think theologically. So, say a Sunday School group is reading about the baptism of Jesus and the teacher, before the class discusses the passage, lays out the following pre-chosen cards creation, mission, and chaos. I can see the members of that group using those cards to more easily enter into a theological discussion of that text at a level that is deeper than where they may have gone during the course of a normal morning.
  • mapping systems. I could see a student reading the first chapter of Augustine's Confessions and using the cards to map out topics as Augustine discusses them. Moreover, by laying out the cards in a tree or a circle or some other relationship and by using the meta-levels encoded into the cards themselves, the reader makes a map of Augustine's treatment. He or she might begin to play around with the organization, asking which are more or less important. And also, by looking at the relationships written on the cards, the reader might discover the third use that I could see this tool providing, namely
  • outlining what's missing. Thanks to the relationships encoded in the language itself, the cards could be used to highlight what is missing in a work that is being read or in one's own thinking. Thus, the language serves to aid and broaden one's own theological thinking.

The degree to which especially the last two aids will work depends largely on how an infinite number of cards may be properly reduced to a fit number of patterns, not ideas. I suppose, if the cards were indexed to their own web pages, as group works has done, any amount of information could be associated with a card. One could, for example, make a list of the theologians who best address a particular pattern and where they do so.

Finally, I need to remember that patterns are responses to problems, where a problem is a common problem in a field. To describe a problem is to describe when a particular pattern is applied. Consider listing potential solutions as well.