Contact:cdent@burningchrome.com
Zachmann, J. A. (1987). A framework for information systems
architecture. _IBM Systems Journal 26_ (3), 276-292.
An apparently seminal work in information architecture that provides a
framework for a common language set for people implementing complex
information systems. A description of the traditional architecture
deliverables is genericized to manufacturing and the mapped to
informations systems. The framework provides a multi-dimensional and
extensible model that can be used by business owners, systems
analysts, systems designers and programmers.
-=-=-
This is fine bit of systems thinking. I heartily approve of modeling
in one domain and mapping the model to another domain. My largest
criticism of this work is that by the end of the paper while I was
conceptually prepared to understand the model the needed recapitulation
was not there. I could not state with any confidence what the
framework is. I assume I'll get that in the next article.
Back to the Index
Contact:cdent@burningchrome.com
Sowa, J. F. & Zachmann, J. A. (1992). Extending and formalizing the
framework for information systems architecture. _IBM Systems Journal
31_ (3), 590-616.
-=-=-
As expected this article does go on to provide a better understanding
of ISA. At the beginning of the article it is summarize succinctly and
displayed graphically in a way that survived the copier.
The authors then formalize the structure by providing rules for the
system. This leads to some interesting thoughts:
Why when creating a model of the universe are the modelers so
insistent on creating rules from the model which they then wish to
apply back to the universe? This sort of mental pigeonholing can be
incredibly limiting and is surprising coming from a paper that
discusses a model that was generated through what I like to call
conceptual overlay: take a concept, map it, try to overlay it in a
different domain, if you get some overlap you may be able to generalize
your model. The model was created by permeating conceptual boundaries.
Rules try to keep boundaries hard.
I'm fairly convinced that Zachman is in this paper so Sowa can use him
for coattails to crow about conceptual graphs.
Rules are inevitable in the application of abstract theories. Especially
when the application has as a goal the codification of the theory in a
computing system.
Conceptual graphs are great for computers but they are too constrained
for people.
I'm impressed that the ISA model is recursive but nor am I surprised:
any model of reality must decompose to allow for the fractal nature of
things.
I think this article demonstrates rather nicely that language
structures thinking and reality: a model of what, how, where was
extended to include who, when, why just because it seemed reasonable,
and it worked...
Back to the Index