After watching Merlin Mann talking about the idea of creativity patterns today, I started reading through volume 16 of Make magazine.  On page 57 there is a brief but interesting article about TRIZ.

TRIZ is an evolving set of patterns for problem solving which was first developed in 1946 by Russian Genrich Altshuller.  It doesn’t appear to be fully available on-line.  In fact, it doesn’t appear to be fully developed into a single definitive framework at all.  Nevertheless it appears to have some interesting things to say about problem solving.

From the Make magazine article:

…Altshuller observed that the same problem types appeared time and time again, and yielded to corresponding generic solutions…

This echos with what Merlin was saying about creativity patterns in his Macworld talk.  But how is this different to just knowledge?  It isn’t, I guess, but these sort of endeavors open up the scope of what we believe it is possible to have organised knowledge about. They also standardise how knowledge is organised.  Standardisation is important (in the sense that it allows efficient communication).

The essence of the TRIZ system appears to be utilising a number of patterns of interventions into a system in order to remove a constraint (or ‘Contradiction’) without compromising the system. To me this feels similar to the generative sequence approach of Christopher Alexander.

Alexander’s generative sequences unfold using a standard set of interventions into structure (which is system-like, I guess).  Also, when the intervention is made you effectively check for ‘compromise’ when you re-evaluate the degree of life / wholeness in the resulting structure before moving on.

This idea of removing ‘contradictions’ in TRIZ  also has echos of all that Ayn Rand ‘check your premises‘ talk.

Interestingly, I have always found the idea that contradictions don’t exist as very helpful in problem solving.  Basically, I see them as a signal to break something up into smaller and / or different parts.  I see disagreements between people as similar types of problems – and the breaking down of concepts into more elementary components usually means agreement can be found.

Of course, it doesn’t really matter if contradictions do or do not exist – the question is whether acting as-if they don’t exist is a useful problem solving tool.  And I think it is.

So, patterns can help problem solving.  And patterns are a ‘just‘ a way of organising knowledge.  And problem solving is about making system or structure interventions and then making sure you haven’t destroyed the integrity of the whole.

Management of knowledge work is a lot like problem solving.  So it’s likely to be pattern-based too.  MWT Collaboration Architectures are patterns…