To: Dave Pollard

Dear Dave

I've just come across your blog, via a link while looking into online tools to support a new 'sustainable communities' initiative I am planning here in London.

I am excited and interested to come across someone who is particularly interested in general approaches to ways for humans to address complexity intelligently.

It is clear to me, as it is to you, that this is urgently necessary, as we have reached a point where tools developed on the basis of radically simplified approaches to reality (using the rational method of addressing only very limited subsets of questions, and rigourously excluding consideration of wider implications ) have become so powerful and widely used, that the inevitable 'unintended consequences' (which occur whenever interventions are made into a truly complex system) threaten to swamp us.



The trouble is, that humans are not built to conceive complexity - our minds cannot hold even trivially complicated ('complicated, knowable' in your classification) subjects in consciousness and analyse implications of changes. When it comes to the truly complex systems, which you rightly recognise include almost all important environments, we are ridiculously poorly equipped.

We need tools to allow our simple minds to rationally address complexity. You are working on various approaches, which I have not yet had time to do more than skim. However, you don't appear to have come across (or if you do, you don't reference, as far as I can find) the work of Christopher Alexander. This is not necessarily surprising - he is primarily interested in, and known for, his work in architecture, but I believe his major achievement is directly relevant to these issues.

The following is a brief attempt to sum up his work in relation to complex systems. If my assumption that you have not come across his books is wrong, the please forgive me.

'A Pattern Language', and various following titles, outlines an approach to architectural design which is not described in terms of complexity. Indeed, published in 1977, it was written before complexity was beginning to be a term that was widely used in any specific sense. However, address complexity is exactly what it does.

The book is structured around 'patterns' - which I like to think of as recurrent eddies in the flow of complex systems - something like the 'strange attractors' of chaos theory - instances of which can form sustainable, vital parts of larger systems, and which support smaller systems. These patterns are linked in a loose hierarchy which is known in mathematical terms as a 'semi-lattice'. It differs from its cousin, the simple 'tree' diagram, in that it allows non-hierarchical linkages, while still capturing parent/child type relationships between elements. The key thing about this is that it is in itself a complex structure -patterns sets have the capacity to map real world complex structures, as they do not demand radical simplification, and are open to continuous refinement.

The links between patterns are made explicit, and the structure of each pattern as set out on paper encourages clear definition of the recurring situation (whether it be a problem or a structure), rational discussion, and crucially, a descriptive, rather than prescriptive, analysis of what is required for successful implementation.

Alexander himself remains committed to architectural practice (a profession which rubs its practitioners' noses in complexity with ruthless regularity!), and has not seemed to appreciate the applicability of this methodology to other complex situations. It has, though, been taken up by computer programmers of a particular bent with astonishing results (indirectly responsible for the amazing phenomenon of the 'wikipedia'). The odd thing about this is that programming is of necessity a practice which uses radical simplification - computers are the apogee of this, in that all problems must finally be reduced to binary questions. Thus discussions of pattern language approaches to non-architectural work have been dominated by programmers, who have not been interested in the power of the approach in relation to complexity.

I am currently involved with setting up a small independent school, and am beginning to use a patterning approach to building a flexible, 'living plan' for the institution, which can embody both ethos and working practice into a single, vital document, which remains part of the everyday life of the school, in a way that 'mission statements' and business plans never seem to.

I am also interested in applying the 'living plan' idea to more straightforward business planning.