Raison d'être
Patterns when applied with energy, adequate social negotiation, critique and sensitivity, represent meta-best practices. They capture the best of the best. An assembly of patterns gives rise to a super language, a high level efficient and very rich discourse. Patterns are part representation, part knowledge artifact (thing), part compact solutions and part validated experience. Patterns may represent strong reification, they carry meaning and an investment of community energy (cathexis) when constructed and shared.
Looking around the knowledge management landscape, I see lots of stuff on intellectual capital, knowledge architectures, data & text mining tools, portals CRM and more. What I do not see is discussions around how knowledge is created, how it is tested, how it gets transformed from implicit personal opinion & experience into shared meaning and becomes part of group understanding.
Most KM folks are busy creating communities, building intranets, selling KM consulting, doing fancy data and text mining, but there is very little going on at the deeper levels. Let's take some examples:
1) How does a group augment and organize its awareness, i.e. practice 'being mindful'?
2) What are the qualities, economic importance and means to validate & improve our collective knowledge?
3) What is the 'organizational mind' , what does thinking together really mean?, how can we improve and use these concepts in business?
4) How can we decode the knowledge & realize economic advantage from knowledge embedded in our collective practices?
5) Can we gather best practices and explore the way a group can know more than the sum of all the individual members knowledge?
6) Is there such a thing as group consciousness?
7) How can we leverage our collective learning?
8) What are effective ways to transfer sticky, situated, distributed knowledge?
Patterns and ontologies may be part of the answer!
What is a Pattern?
Each pattern defines a context of use, a recurrent problem that needs to solved, a group of "forces" pushing and pulling in different directions, and how those forces might be resolved to best solve the problem - a working solution based on validated experience. Both good and bad (anti-patterns) examples are also provided.
Pattern Language is an attempt to analyze, group, store and reuse ageless wisdom. They are not so abstract as to require the rediscovery of how to apply them successfully, nor are they overly specific to a certain situation or culture. They are somewhere in between. A pattern explicitly describes possible solutions to a issues within a specified or similar context, by describing the qualities of working solutions and the steps to implement them.
Remember that a pattern language does not only serve as a guideline for building or designing something, but as a system for talking about, deconstructing and classifying things.
Read further:
http://www.cpsr.org/conferences/diac02/patterns/patterns.html
I will cover ontologies later
Calling things best practices drives me nuts. They are practices. We won't know if they are best practices until we get down the road a bit.
Pattern languages are interesting, because Chris Alexander is always referenced. In one of his later books, the one where they build a model city using his pattern language, he didn't like the results. After that he moved on to emergence.
The software world only adopted pattern languages after Alexander had abandoned them.
Still frames, augmented transition networks, even heuristics evolve around recognizing patterns. They were just different ways to do so. That they weren't human readable or sociable is why they are representations and stop with that.
That knowledge doesn't have to be created or explicated is amazing to me. That tacit can arise without human intervention and be conveyed to others without explication is the interesting thing for me.
Posted by: David Locke | October 08, 2003 at 03:26 AM
denham... i posted a reference on my weblog this morning to the cutter consortium's september journal which has a focus on pattern languages with articles by: Jim Coplien, Alan Shalloway, Steve Berczuk, Dirk Riehle, and Linda Rising and Esther Derby. unfortunately, this is a "for pay" journal but the introduction is freely available for viewing... judith
Posted by: judith | September 23, 2003 at 08:38 AM