John Zachman originally called row one of his framework, “Scope” (1), representing the point of view of the CEO who is looking at an enterprise as a whole. One of the rules of the framework,
however, is that in moving from row to row, you are not simply changing level of detail. For this reason, it isn’t really appropriate to imagine that scope is simply the “Business View” at a
higher level. What does that mean “scope” really is? How is it fundamentally different from row two’s “Business View”?
Simultaneous with the development and propagation of the Zachman Framework, there has been a lot of work in the development of “patterns” in system architecture. Your author’s Data Model
Patterns: Conventions of Thought (2)attempts to present standard data models for standard situations. Analysis Patterns (3) by Martin Fowler does something similar, with an object
orientation. Design Patterns (4) by Erich Gamma, et al. presents object program module designs that can be reused.
Preceding all of this, however, is the work in 1979 of Christopher Alexander, an architect from Berkley, California. His book The Timeless Way of Building (5) describes an approach to
architecture (physical buildings, that is ¾ not system architecture) which acknowledges that we all use patterns when we build things. The trick, he says, is to be sure that you use patterns
that have “the quality without a name”. This quality, which buildings and cities can have is “the difference between health and sickness, wholeness and dividedness, self-maintenance and self-destruction.” There is a difference between patterns that have this quality and those that do not. “In a world that is healthy, whole, alive, and self-maintaining, people themselves can be alive and self-creating. In a world which is un-whole and self-destroying, people cannot be alive. They will inevitably themselves be self-destroying, and miserable. It happens that the single central quality which makes the difference cannot be named.” (6)
Too many of the patterns we use ¾ in buildings and in systems ¾ do not have that quality.
To describe the quality, Christopher tries to use the words “alive”, “whole”, “comfortable”, “free”, “exact”, “egoless”, and “eternal”. But in each case, the word meant more than the
essence he was trying to get at. He did say that perception of the quality is not subjective. We all recognize it when we see it. It has simply been in rather short supply in the architecture of
the late twentieth century.
His patterns are hierarchical, with each pattern intimately tied to those above it and below it. He first addresses whole regions, with patterns for “The distribution of towns”, “City country fingers”, “The countryside”, and so forth. He then describes major city structures as being concerned with “Mosaic of subcultures”, “Local transport areas”, and so forth. From here, he goes through various levels of communities until he gets to neighborhoods that may use such patterns as “Activity nodes”, a “Promenade”, or “Night life”. When you get to the level of houses
themselves, you have patterns for “The family”, a “House for a small family”, “House for one person”, etc. An office building is made up of the patterns “Number of stories”, “circulation realms”, “Family of entrances”, among others. Finally, within one’s house, the back yard may use the patterns “Terraced slope”, “Tree places”, or “Trellised walk”. The house itself is a collection of such patterns as “Window place”, “Eating atmosphere”, “Bed alcove”, and so forth.
The entire set of patterns are presented in a second book, A Pattern Language (7), published in 1977.
What is interesting about Christopher’s work is not so much the patterns themselves (although you now have the opportunity to understand why most office buildings are so awful), but his approach
to the use of patterns. In recent times, we have been carried away with the idea of “modularity”, as though we could build our world (and systems) from a set of Lego blocks. But the fact of the
matter is that the toys built with Lego blocks tend to look stilted and artificial. The apartments and subdivisions that are a single structure repeated endlessly are incredibly banal. Better, says
Christopher, to have a set of patterns that describe the essence of the elements to be used in building. Then what you do is have the people who will live in the buildings to specify exactly how
the patterns should be implemented to give them something unique and special. Each person then will have adapted the patterns to his own personality and interests.
Approached this way, you get infinite variety, even though all the houses and buildings come out with similar character.
So what does this have to do with building systems?
When you set out to create a business, you have something in mind. You know about other businesses in the same industry. You know the rules of bookkeeping (well some of us do), the tax laws, and
the other logistical constraints that you must work within. Even if you are creating an entirely new business (think Amazon.com or Federal Express), you have patterns in your head which will guide
you. You are simply applying them in a new way.
Even though John didn’t recognize this when he laid it out, this turns out to be the domain of row one of the Zachman Framework.
Think about it: In the “data” column, we can begin with patterns of data organization. We know, before the company is set up that it will require data describing people and organizations,
products, contracts, and so forth. In the “process” column, if we are manufacturing something, we know that we will have to transform raw materials into finished products, according to patterns
that are derived from the nature of the products being manufactured. In service industries, our processes are derived from patterns for the services to be provided. There are patterns for doing
these things that have, in some cases, been laid out for many hundreds of years ¾ although technology has had a dramatic effect on many of these patterns in recent years.
In the distribution of our business, there are patterns for distribution and logistics. One way or another, we will have to acquire raw materials (often from distant places), and we will have to
get our products or services to the market. In addition, there are patterns for the organization of headquarters, branch offices, and the like. In the “people” column, there are patterns for the
organization of companies, and in the “timing” column, there are patterns for the sequence of planning and execution. And there are standard patterns for business rules in the “motivation”
So, the models that constitute row one will be models of these patterns. The assignment is to produce “good” patterns ¾ those which have the “quality without a name”.
Note that these patterns precede the business view, and therefore considerably precede any system view. These are patterns for the organization of the business itself. Once we have identified these
patterns, we can then also use them to guide the design of systems to support that business.
I can speak only from the experience of addressing data patterns in my book. It is the case that when I have presented models based on these patterns to business executives, they recognized them
immediately. These patterns described something fundamental to the business that went way beyond the particular events and objects perceived by people working in the plant or office. It is also the
case that, while I use the same patterns to develop models for all my clients (be they in clinical research, oil refining, news gathering, cable television, or manufacturing) they are all different
in subtle ways. Each company has things that are more important, and its own particular slant on how to do things.
Recognizing that Row One of the Zachman Framework is patterns provides us with a powerful tool not only in the development of a system, but in the development of the business it will support in the
first place. The trick here is to find the right patterns so that the business you build will have “the quality with no name”. Herein we have the true mission of business process re-engineering
¾ and a way to approach it.
- Zachman, John,
- David C. Hay, Data Model Patterns: Conventions of Thought
- Martin Fowler, Analysis Patterns
- Erich Gamma, Richard Helm, Ralph Johnson, John Vlissides, Design Patterns
- Alexander Christopher, The Timeless Way of Building
- Ibid, Page 25.
- ______, A Pattern Language