• An honours thesis in human ecology – see that sub-project page
• blog
• my reading list

What is statistics? To consider: What happens when we come to say that we know the probability of a die landing on a given face is 1/6? it means that we have sufficiently many sufficiently unknown variables in the throwing of the die that are sufficiently randomly correlated with the eventual value of the desired unknown that we can regard the distribution as some kind of nondeterministic well-behaved variable?

For functions that are less well de-coupled from their variables, um, what can statistics mean? SAy, the probability of destructive storms, which is coupled with a lot of variables, say, human response to storms, which affect the parameters?

For the purposes of this paragraph, I come from a background of practical political engagement with policy on environmental issues, and combined degrees in human ecology and mathematics. I am experimenting with the idea that by stepping back from practice and considering theory I will find new ways of being practically effective. Ideas to toy with:

• I want to step a long way back. If trying to unify statistical mechanics with ecosystems metabolism plus a dash of Foucauldian geneology of ideas will shine new light on public policy, I am happy to go there. (Steps to a phylogenetics of ideology?)
• speaking of which, I have vague suspicions that I would like to raid cultural theory for insights on how to use hard science tools, such as systems modelling
• I would like to play with ideas about science as a knowledge system which is embedded in certain cultural problem-solving systems. Maybe it is not great in providing answers to resource management questions because it has not been created as a system to solve problems in that domain. If that is so, what sort of knowledge systems are apropriate to solve such questions? (Compare and contrast agile processes in software development and business management, and adaptive management in participatory resource management. Further: considering the knowledge dissemination systems which encapsulate knowledge generation systems. It’s great coming up with a trendy new way of analysing the world, but how to you get it out into the current global mediasphere?
• sometimes I look at recent results in quantitive biology or ecosystems metabolism or systems thinking and get an inkling that I am seeing a glimpse of the arhitecure of life, and I just have to pursue it.
• This wiki is an experiment in tying these ideas together. I do want to keep it unabashedly academic, though. I’m not into the internecine disputes about base principles that characterise Wikipedia. But how do you maintain such a thing?
• Note to self: Demolish the presumption of intensionality.

I’m ruminating on:

• distributed cognition and the impact that it might have on theories of language, organisational complexity, and bunches of random stuff like innovation theory and social capital.
• discretisation in biological models and how to do it.
• learning as a thermodynamic process
• systems of learning as institutions themselves subject to feedback
• Gerd Gigerenzer and his thoughts about heuristics
• explaining this wiki and getting people involved
• decision making in ecosystems, human systems and bacteria
• generalisation, as an evolutionary process, as a scientific method... about the concept of a statistical ensemble and novel ways of getting on in unrepeatable experiments... about science on the edge of repeatability and reproducibility Is the scientific method, in essence, isolating structures that are reproducible? In which case, how can we find reproducible structures underpinning unique complex systems?
• What is Complexity?
• What is systems thinking? and why? We think about interactions with our environments in terms of systems why? To flag that they might experience feedback dynamics? simple interactions will be embedded in wider systems dynamics?
• cross-scale interactions. What is is that it seems that order is traded up and down the hierachices of scale of interacting complex systems... the fast and loose networks of human societies are made up of individuals composed of tighly organised organs, cells and brains... free market economies are made up of tighly controlled firms and households (?)... in the opposite direction, large manufacturing processes dominate over simple cottage industries, and human beings displace cognitively sophisticated creatures of no exploitation value to them. Actually, now I think about it, these examples demonstrate a tendency for larger-scale systems to be more ‘loose’ (whatever that means) than smaller ones, and for more complex, aggregated systmes to displace less complex ones at the same hierarchy (in some sense) within the wider ecosystem.
• Thinking about how academic publishing includes a model of veracity, in its peer-review and citation protocols. but it doesn’t allow error bounds, sensitivity analysis, of veracity.
• How is software development like normal science? Or post-normal science? Model development? Need to include model developer within model? Iterative process? Does normal science also have a process of model development, but one that home in on the bits of the world approximating linearity? What domains do these processes constrain themselves to? agile proccesses in Model development?
• energy in ecosystems and its limits as a tool for understanding
• nested rationalities - i’ve already forgotten what i meant by this, but i think it was something like what Minsky would have called frames.
• computers - why do they change things? Why isn’t information already processed like they do it?
• “bricolage” as an alternative understanding of conitive modelling, as opp metaphor
• actors vs context
• what is mathematical information- how does it relate to economic information or ecological information?
• what isn’t a surrogate variable?
• variables in systems models

Ontologies of complex dynamical systems? Radial categories versus fuzzy sets in systems thinking? The thermodynamics of Model formation? agile processes versus post normal science? metabolism of information systems? Evolutionary rationality? Ecosystems as syntactic creatures?

• There’s a lot of grandiose concepts i want to play with, with a lot of rteading behind them. I need to find a question which is rich enough to fit the contraints of an honours thesis and still give me an excuse to read all those pontificatory texts.
• Want a systems dynamics thing so as to maximise the benefit of having Barry Newell around and prepared to consult. (But i am interested in the transformation stage, the restructuring of a system in crisis, he seems more intent on the fairly stable intermediate stage wherein the system manufactures its crisis, but not the actual crisis itself.)

Ontology in an computational science sense, more than a foraml philosophical one. How to factor a system into objects, and why. How to refactor. What this says about the dynamics. Lakoffian notions of models. when is a factorisation defunct?

this implies questions about the purpose of dynamical systems thinking.

• Q: What do we do it for?
• A: To identify elements of surprise that linear thinking would not uncover. So we want a lower and upper bound on its utility.

yeah, i like this one because it has good scope for exploring

• the goal of dynamical systems modelling
• the models used by agents within the systems
• the cognitive process of modelling
• what the evoloutionary actors that we might suppose are acting to produce meaningful factorisation
• methodologies of modelling in general
• what reproducibility means across non-replicatable systems (what abstractions may be said to be reproducible?)
• in what sense we are aproximating when we make these models. (sensitivity analysis of fuzzy modelling?)
• how the hierarchy of panarchy interact when tied across scales and regions - as in barry newell’s models where different regional effects become tied in
• what sort of objects we can expect to be the subsystems and actors in different regimes.

And as a random idea

• rewriting systems models into layers of effects of different timescale.

• What this question shuts out that I would like to explore: Tainter et al’s “Resource Transitions and Energy Gain: Contexts of Organization” talks about the types of dynamics that are likely to donimate in different energy contexts. I’d love to have a crack at that - it seems to apprach things from the other end very nicely.
• Where is the room in all this to play?
• what is my (meta-)methodology for all this?
• ecology has given me all these interest and insights, and it is no longer core to my research program. I feel an emptiness, where once lived my superior and snide inner ecologist.

• the pheonix people’s study of 100 cities urban dynamics- raises questions of how to aggregate information - when do a bunch of events form a class? how do you select a bunch of ecosystems surprise events and identify them as belonging to the same set? This relates to Cosma’s Shalizi'a observation that examining the world can be likened to a process of inferring a grammar, or a hidden markov model.
• comp-sci metamodelling
• analysing the domain over which a model produces useful predictions (the opposite of sensitivity analysis?) is a useful goal for systems
• who the hell will mark this kind of paper?

• produce a decision tree to choose granularity of objects in a mdoel for a given problem?
• produce a guide to degree of aggregation in the, shall we call them, hidden markov states?