Complex Systems Science in the Urban Context
From COSNet
This page is for continuing interactions from the 1st workshop of the Complex Dynamics of Urban Systems Project.
On the 3rd and 4th of October 2007 CSIRO hosted the first in a series of workshops funded through a cross-divisional project emphasising the potential for applying CSS in the urban context. The objective was to bring together researchers working at the interface of urban planning and management and those with expertise in the application of complex systems science approaches to interact and share knowledge. Participants were:
- Researchers in urban studies, planning and sustainability
- Modellers of cities and urban systems
- Researchers in complex systems science, particularly those with experience or an interest in applying it in the urban context.
The environment of the workshop was sufficiently informal that participants felt like they could ask both the ‘big questions’ and the ‘dumb questions’ about each others’ area of research. This interaction has proved to be a fertile way of generating a broad understanding of the potential for applying complex systems science in the urban context.
Presentations and discussions at the workshop included:
- Identifying major urban issues: water, energy, urban development, “resilience”, that could usefully be addressed with complex systems science. What urban issues are not complex?
- Examples of current urban research (not necessarily complex systems based) e.g. decision support tools, urban metabolism studies.
- Where could complex techniques usefully interact with that urban research?
- What complex methods or approaches exist? How can they provide insight into urban dynamics?
back to Complex Dynamics of Urban Systems
Day 1
How to model urban dynamics, transition, evolution? How and why cities change and what are the best modelling approaches?
- Why cities change? Cities as attractors: opportunity of generating economic value e.g for the entrepreneur. Note also that negative attributes might trigger positive change e.g. taking the traffic out of the centre of Sydney as a result of over congestion may improve the residential qualities of the city.
- "Locking in factors": (inadvertantly) maintaining the staus quo e.g. building new freeways encourages people to use cars more and then we look at car usage and think that we need to prioritise builidng and maintaining roads.
- Pioneers or champions are necessary for "unlocking" - bottom up approach has limitations - city scale leadership and capability
- What's the necessary condition for a champion to emerge: crisis and the social conditions for the champion to act as a champion
- What's the non-crisis, non-negative factor that causes major change in cities? technology and invention, externalities to the cities/state, globalization.
- Sometimes the technology is available but the social system is not responsive to the positive opportunity
- How to model a system that crosses a threshold and yet survives or even thrives.
- Valuation of existing greenspace as much as development
Future urban water systems design - the impact of complex systems
- integration with other systems
- impact of new technologies and approaches on restrictions
- lifetime reponsibility for operation and maintenance of 'future system'
- introduce reliability, robustness into design approach.
City service interrelationships
- Resilience of cities can be used to allow the interdependences to be assessed: what is resilience?
- Indicators or measures to gauge resilience need to be defined.
- What is the realtionship between resilience, sustainability and liveability?
- How do you explore interdependencies? Basic data sets needed.
- Need to consider interrelationships between different areas and indentify relationships between data in each domain and the implications of change
- Tolerance to accept different service levels
- Community involvement
- Suggested that potential exists for collaboration
Nesting (complex) agricultural systems in urban systems
- Analogy of an operation which isolates the essential part of the system in order to identify processes that are critical
to the systems of immediate concern
- Subsuming out-of-system inputs as some cloud, as long as they are understood and explained well, can be a valid approach.
- There must be a progressive embellishment of the feedback between the systems
Pictures of the complex city
- Showing ALL links for legislation and responsibility leads to a 'horendogram'. Reduction by role showed who mattered and where people or organisations may not be linked to the decision process.
- Dynamic representation of resilience - attractor basins linked to networks and binary - emphasises that you can look at dynamics with out precise simulation and mathematics. You can identify the POTENTIAL for thresholds and help identify real thresholds.
- We tend to worry to much about accuracy and lose the key message by swamping it in detail
- Can use size of component to represent it importance
- Need to simplify connections where one system needs input from a more complex larger system
- A limited number of simulations can provide appropriate detail - see pictures
Setting objectives and measuring success across the city
- Containing objectives to appropriate temporal and spatial scales, ideal vs pragmatic - boundaries are often defined by history, geography, culture and other data.
- Objectives are emergent - action learning and research
- Balance between measures that can be quantified to show success to stakeholders and qualitative measure which have an inherent stochastic variation.
- Need to be realistic to ensure outcomes. So limit scope of objectives to a subset of reality that CAN be "simulated".
- We are working towards a society that has improved reachenalety with respect to the environment. Complex systems science allows us to think bigger, wider and set more reaslistic/fundamental objectives.
Validation in Complex Systems Science 1
- Complex systems exhibit feedback, non-linearities and sensitivity to initial conditions (chaotic behaviour)
- Validation is comparing what exists in the real world and asking: does my model replicate that? Just one problem, the intitial conditions may be replicated but the model diverges - need to make sure the model reflects the ATTRACTOR
- data assimilation - live feedback in the model
- getting dynamic information back from people as opposed to static e.g. filling in a survey - is this made more possible with interactive communications and the internet.
- Inverse modelling: predict backwards to find initial conditions
- Probability density function estimation
- Assumptions greatly affect the construction of the model
- How do things relate to one another: feedback loops
- Statistical analysis of dynamics models
Validation in Complex Systems Science 2 - The role of agents in complex systems modelling
- Agent based models (ABM) tend to only work in the small scale
- Big models mean a big number of nodes and a big number of interactions and a big number of intervention points
- Social network analysis - criteria, analyse network for metrics such as centrality and intervention e.g. a disturbing spread of HIV in a community
- Maximum Likelihood Markov Chain Estimation - attempting to fit network structures to what you see in reality e.g. Lampert et al. predicting the next 100 years through samples of simulations
- One way of looking at this is to ask: what conditions will make your model fail? Construct models that have a lesser number of points that allow it to collapse. Focus on robust models
- What conditions would not allow my assumptions to hold
- Predictive or Descriptive models - interdependencies between variables
- Risk - compounding effects of variables /parameters
- Machine learning has the STAT3 test, we need a generic way to do testing/sensitivity analysis: multiple validation tests are needed
- Best Practice Approach: is there a standard model to compare against? Compare against the "best" model e.g. epidemiological models such as EPIDIMUS. Where are the large degrees of freedom in terms of results?
- Should we design tests at the requirements stage?
- Calibration of Parameters - verification e.g. debugging, validation e.g. structural testing.
Feeding a city in the face of Peak Oil and peri-urban development
- local growing could become competitive if the price of oil rises such that transporting food into a city (or competition for the food from outside the city) causes prices to rise. Note that already there is $1 billion worth of agricultural production in the Sydney Basin.
- Timescales: urban development seems more urgent now but the low cost of oil now means that it is part of a positive feedback that allows cities to expand outwards. In the event of sustained rises in oil prices, the cost of transport fuels will act as a negative feedback to constrain the outward growth of residential development.
- Want to demonstrate that encouraging e.g. more food self-sufficiency in the city contributes to multiple benevolent feedbacks: fresher food, shorter supply chains, re-connection of the people who supply to the city with the people of the city. There may also be advantages of using water where most of the rain falls though it should be remembered that a city's inhabitants will need substantial amounts of water for residential purposes (possibly more in higher density circumstances) and this may present as competition with water for urban agriculture.
- The people who currently own the peri-urban agricultural areas are relying on selling the farm to provide for their retirement. This exposes the good quality agricultural land to development because it has higher value if developed. How to maintain land tenure, continuity in land use - financial issues. How to avoid "adaptive cul-de-sacs": regulations and legislation that mean a "dead end" even for those who are willing to be innovative.
- Urban agriculture has been established successfully in Havana, Cuba in response to the twin stresses of the Soviet Bloc collapse and further trade embargoes imposed by U.S.A. This had the effect of cutting their fuel supplies in half (in effect a mini-peak oil crisis).
- Glocalisation: the local response to a global event or continuing action.
- One way to seed the positive feedback for urban agriculture is engaging the local community to be aware and even take pride in the local produce. (see also the Slow Food Movement)
- Some parameters of a possible study to e.g. look at feasibility of a "Food Secure Sydney" maybe to compare with other cities?
$price of oil $ cost per ton* km of transporting produce (logistics companies may allow access to aggregate information) $ cost of delays due to traffic congestion? $ cost of road repairs? trucks cause several orders of magnitude more damage to roads than cars - what if there were more trucks with more local agriculture? land use and soil quality maps urban garden space water access proximity of production to demand availability of fertilizer (see also mining the waste of the city) travel time contours (connected with development, mass transit has more stable contour structure compared with road transport) finacial incentives or regulations to keep productive land in or around the city. rate of obesity in the population nutrient needs of the city population (total tons and /capita) capacity of governance to act will of community (allowing time to consult and setting the system boundaries)
Mining the city - eco-industrial review
- We lack the data on specific information on the spatial location of waste and potential materials which can be recovered from cities
ref to Graedel
- We need to map the relationships of firms and industries which create opportunities for synergies in waste exchange and the benefits of co-location
ref to I.E discussion point
- How do we go about materials recovery from the built environment (circular economy in Japan)
- It's relatively easy to evaluate the potential for utility exchange but modelling the potential of by-products is difficult.
- Research in this area is fragmented, how can we develop a better network of research and exchange data.
- Optimise the location of waste using industries to create opportunities for industrial ecology
- Need to look at gaps in knowledge and information
- How can we marry the concept of industrial clusters with industrial ecology?
- Need to develop an Australian Industrial Ecology Network.
Complex models have to be used to model complex systems
- Complex models DON'T have to be used for complex systems. Models are an abstraction of reality checked/validated and updated by reality.
- The most important thing is to select a proper model to meet specific target objectives while satisfying certain criteria (resolution, scale, boundaries, values, cost etc.) rather than simple justifications for simple or complex models.
- It's important to have meta models
Urban systems as part of the natural environment
- Cities exist in a context including surrounding natural environments
- Although humans draw boundaries between urban and peri-urban or conservation areas, these boundaries are not real or physical (they are a model); they represent a reductionist approach that limits our ability to manage the sustainability of both the natural areas and the urban areas that depend upon them.
- For instance we are currently trying to manage weeds in the wet tropics World Heritage area near Cairns. We go out into the rainforest to manually remove and manage weeds but we start the arduous physical search at the nurseries that are the source of the origianl infestation. These nurseries are still operating, bringing in the weeds that we will still be managing in 10 years. Unfortunately these nurseries are preferentially at the edge of the urban-rainforest interface and we need to capture that information to effectively manage weeds in to the future.
- We need to assess the importance of ecology to our city life as well as the impact of city life on ecological systems. If ecology IS important, ecological indicators could be a measure of urban sustainability.
Bottom up approach of modelling urban complex systems (production and household interactions)
- The bottom up approach: Households/factories/enterprises ->precinct ->council ->city ->state or nation
- At the national leve, aggregation of data converges to systems of national accounts, physical accounts, stocks and flows
- At the bottom and intermediate level it's the interaction between agents of production, industries and households.
- The suggestion is that starting with a description (account?) of these lower level physical transactions, you aggregate into national accounts of $ and physical stocks and flows
- The importance of putting in local factors.
COMMENT There are issues of data availability here - e.g. although there are household consumption surveys, they are not for every household in the city - at some stage an inference may have to be made that the attributes of a sample represent those of the whole. This can reduce the problem to one of "disorganised complexity" where averages and statistical distributions apply to a population and you lose (or don't need) information about individual interactions.
- Issues: lower level agents do not equal emergent behaviour
Capacity Tensions
- Do the solutions respond to prescriptive ideals or descriptive understanding of the urban system? Do the outcomes reflect badly on systems theory or our ability to use it?
- It's important to remember that systems are 'lumpy'. An improvement to an infrastrucutre network is spasmodic while demand keeps growing over time.
- Drivers like peak oil and climate change could be reasons to address capacity tensions or we can conceptualise the idea of capacity tensions
- Capacity tensions can be addressed in a variety of ways, not just the supply of additional infrastructure capacity
- How do you decide WHEN is the right time to introduce a new element of or upgrade to infrastructure? CSS might help answer that question.
- Analysis using CSS using CSS is expensive which is why it is not often used
Concepts of Resilience - identification and measurement of resilience
Three types of resilience are identified:
- resilience against regime change
- resilience for response and recovery
- resilience for adaptive capacity of systems
- In social systems, governance is important but hard to express and even quantify with respect to resilience
- A theory of "cummulative causation" developed around the 60's (George Galstar, Gunnar Myrdal) which could offer some inspiration for resilience measurement
- Resilience could be dealt with in a similar fashion as risk, and both be treated in tandem in a systematic way
- The experience of the insurance industry could also be a source of reference for resilience measurement.
- Resilience is a property a system has and what exogenic forces shape it.
COMMENT I think of the concept of resilience in the same way I think of the concept of "automation" i.e. you can readily see that something is more or less automated though it is difficult to quantify because this may occur in a multitude of different ways. There is perhaps no such thing as "fully automated" but there's certainly a gradient in the degree to which something is automated. Possibly the more useful thing to identify is the quality of being automated, say, compared to being entirely manual. Translating back into resilience-speak: a resilient system will be able respond and recover, it will have a certain robustness to it and yet be adaptable in the event of that robustness being overcome. The degree to which systems differ in their resilience is not as important as the complete absence of the components of resilience. Tim Baynes 11-10-07
Emerging domestic energy patterns from the interaction with other urban issues (facilities, transportation, income distribution)
Relationship between urban density and sustainability
Urban density of Australian cities is falling as cities are spreading at rates faster than population growth. This is adding to the cost of building and running cities. How do we develop cities that balance the need for greenfield development vs urban development?
- How can we develop a better pattern of urban development which makes it more conducive to supporting public transport systems?
- How do we develop a more integrated approach to land use, transportation and employment areas?
- How do we change the pattern of urban development to generate more sustainable development outcomes especially a move to the greater use of public transport.
- We are consuming the most productive land for urban development forcing food producers onto more marginal land which is less productive
COMMENT: More may be different - cities may not just be attracting bigger costs in proportion to the size of the city. Can investment in good public transport induce a change in the residential development market i.e. if an area is serviced better by public transport are you more willing to live there even though it's higher density than the more spacey outer suburbs where you need a car? Note that there was a well thought out submission from the International Public Transport Associationrecently to the Australian Federal Government inquiry into a Sustainability Charter. Refer also to the discussion on "Feeding a City" Tim Baynes 11-10-07
- Low density urban fringe development costs for housing is low compared to inner city renewal areas. However operational costs of low density development are much greater than medium density areas.
How will climate change impact future city development?
Increasing emissions - a climatic opportunity
Playing the devils advocate...
- 1. Emissions can spawn new industries and therefore increase the GDP of a country. Emissions can be captured and used by industries as raw materials. infrastructure needs to be built, more jobs, more emissions, more raw materials. Is the net impact a reduction in emissions?
- 2. From an optimal decision point of view, is it wise for countries like Australia to reduce emissions when its impact will be small when resources (e.g. finance) are limited. Operations research suggests use the resources in the most efficient way with constraints. Therefore, are there opportunities for countries that have low impact to the global emissions to increase emissions so as to displace emissions of high impact countries and achieve a globally optimal emissions reduction. How could such as scheme operate (e.g by imposing a tax on imports from high impact coutries)?
- What are the complexities of the system? can it be modelled in ana ecomnomic system without discriminating against particular approaches
- Some participants recognise thas as an approach for consideration and modelling. Others were highly unaccepting of the idea raising objections ranging from the moral, political and social to arguments that this was counter-intuituve to game theory. etc.
- Can a complex system model be developed to study the impact of such a global approach? Can an economic and politically correct global approach (other than Kyoto agreements) be used to divide responsibility across countries that emit and countries that consume goods from emitting countries> Is per capita emissions a good indicator?
Day 2
Data what do we need? Where do we get it? How do we share it?
- ABS is the biggest data source though CSIRO has collected a lot of data over the years and may be a source for specialised information. Issues of proprietary information and cost.
- Generally knowing about data and who has it, who is collecting it: lots of personal relationships and connections and ad hoc data
- Other, more technical, factors: data assimilation, integration, data quality
- Longitudinal data, data storage
- There's no easy fix and current ad hoc relationships with providers will continue
COMMENT - perhaps this is the ONLY way things can continue because there's an infinity of possible data requirements that obviously ABS and others can not provide for. Formal data acquisiation and sharing may be accessible and accountable but it will be limited. Informally supplied data may not be so easy to find or verify but may be closer to what you need. May be we need a data equivalent of EBay - a clearing house of who has what... - Tim Baynes 19-10-07
Stocks and Flows: issues in integration with energy sector modelling
- The current state of the NEMSIM project looks at electricity market modelling for medium to long term predictive outcomes. The directions for the project have been to integrate other energy sectors. Possible future directions can include stocks and flows to get a long-term predictive outcome
- The idea is to collaborate with work done by other people and include their findings in the NEMSIM model
COMMENT - The Australian Stocks and Flows Framework (ASFF)is not so much a single model as a framework that integrates sectoral sub-models. These sub-models represent the dynamics and transactions of physical stocks (such as: cars, houses, people) and flows (such as: fuel consumption, waste from demolition, effluent and emissions) that describe sectors of society. While the design of these sub-models and their interconnecting structures permit sophisticated future scenarios, information in ASFF flows in one direction. That is to say, many drivers may influence outputs in compound and complicated ways but not involving ‘complex’ feedbacks. This makes the outputs of the framework explicit and tractable but it also allows the framework to be open to input from other models, analyses and expert knowledge. This makes ASFF candidate for interacting with other models that DO provide the complex feedback information (I think such an exercise is being sugggested here).
ASFF's breadth of view in representing many sectors is matched by the depth of historical data (many decades before the starting year of simulations)incorporated into . With this ASFF is aware of the lifetime and age structure of capital stocks such as electricity generation plant. Knowing when this plant will need to be retired enables us to anticipate opportunities for alternative replacements and the indirect impacts of this choice on other, seemingly remote sectors e.g. the water sector.
For more information about stocks and flows frameworks please get in touch with Tim Baynes, CSIRO who is also the contact for this wiki site
How to go about modelling social systems? Is evolutionary game theory the answer?
- Synchronisation of behaviour in groups driven by feedback (ref to Strogatz)
- what about top-down dynamics such as community leaders within the social group
- Example from agriculture in Murdoch, W.A - how are decisions made in catchment areas? partly using census data (ref) has sketched out the network of connections between farmers as well as social norms and policies. Myers-Briggs tests were used to analyse natural leaders, followers and communicators and personality types in general. It was found that the best way to convey new ideas into the community was through farmers wives.
- intervening scientists: are they being listened to? Are they the best people to intervene?
- Can we program Myers-Briggs personality types into an ABM?
COMMENT- while we might be able to simulate the reaction to stimulus this way, it's a subtley different matter to determine reponse
- What is the level of abstraction?
- what is the level of aggregation?
- what level of granularity do you want to view the problem from?
- Granularity - aggregating to the city level, ABM can inform but not really influence communities - engagemnet is hard
- Agents representing individual, groups and stakeholders - game theory in community based models.
- Stakeholder engagement - There's a missed component but have systems become sufficiently complicated?
- Behavioural patterns <-> impact on environment
- Soft systems - better as a community model
- Assumptions are a huge factor in performance - where have previous things gone wrong?
- Structural change in society seems to come about through the influence of a crisis or a pioneer - how can we identify those who can influence a community before there is a crisis event? Can we help them influence others?
- Is the decision process of groups done through leaders or people of a particular personality type?
How do you define change in a changing system - social, economic, environmental, cultural?
- Signal to Noise Ratio
- What do you do with the information: How relelvant is it?
- Pre-emptive action before change occurs?
- Change in which subsystem(s) - soft and/or hard urban systems
- Changes in population dynamics
- Changes in causes versus performance of the system
- Overtime systems iterations
- At what point does change affect performance
- Collaborative structures
- Tendency to measure impacts from an abstract notion without figuring out complexity
- Like discrete systems but weather has patterns
- What is business as usual for cities?
- Distinguish between endogenous and exogenous change in cities
- interpolation vs extrapolation - relationships
- Impact to predictability?
- Individual behaviours vs system behaviours
- You don't know how people will change in advance(?)
- Synthesized model scales
- Stochasticity and randomness - what is the role?
- We don't incorporate political issues
- Decision and risk assessment
- What are the half dozen (or so) properties that will tell us the large part of the uncertainty? (sorts of indicators and their dynamics)
- gini coeeficient
- single indicators
- environmental indicators
- happiness indicators
- a Star rating for cities
- sustainability
- rates
- network characteristics
- Scale and Segregations
- Liveability index
- Cultural differences
- Health care measures (e.g adopted in current economics)
- Social sustainability and hidden networks of care
- Idea to take further and have people evaluating it
- Slow versus fast variables
- Who do we work with and how do we do this eg. to be comprehensive and be able to replicate the work etc.
- Scientists as academics in cities
- Reversing the question: what are the minimum interventions that can manage change given that there are finite resources, time and effort available?
How to integrate best practice and failure examples into urban management through enhancing institutional memories?
- Good way to build capacity/ make up capacity lost.
- Management becomes to detached from the field they are managing
- Cased based approaches
- Problem solving processes might be changing over time - important to incorporate processes into the system
- importance of clear criteria/indexing system
- set up a ratings system of best practices
- LGP, IWA potential clients
- Science component: optimization processes, analysis of best practices, empirical evidence collection
- Good application potential in developing cities
Automating the stakeholder - useful ways of integrating ABM and other complex models with urban metabolism frameworks.
The premise behind this session was to explore the notion of simulating one or more stakeholders in a situation where multiple stakeholders might interact possibly over different time scales. The case of the stocks and flows frameworks is an example of where, hypothetically, it would be ideal to get stakeholders representing all local governments, utilities, residential groups, state government departments, planners and industries to repond to simulated tensions calculated in the scenarios generated using the stocks and flows frameworks. That such a gathering is highly unlikely suggests there's a need to 'automate' one or more of the stakeholders so that you can effectively represent their input while dealing directly with a limited subset of users/participants.
So how do we represent the stakeholder?
- Alternative paradigms: e.g. dynamic systems model may represent the behaviour sufficiently well OR you can identify some major features of the structure of the system. The idea is that independently of how the decisions are being made, if they are being made at an inappropriate time scales, then this can be discovered with a dynamics systems model.
- Can perhaps use dynamic systems model in concert with agent based model (ABM) for more "accurate" practical purposes
- Another paradigm is using neural nets but you need a lot of data to train them and they can only be used in the limited range over which it has been trained. Not adaptable
- Bayesian networks also need data to train them but, unlike neural networks, they are quite accessible and you could show them to stakeholders. They can be dynamic but they're not like ABM because the connections between nodes are explicit
Modelling "fit for purpose" - when CSS will help and what models and decisions are appropriate at what spatio-temporal scales?
- CSS good for largish systems 10 - 1000 agents
- Link to engineering decisions about scale and density
- governance structures are a problem
- Path dependence, good decisions remembered and the lingering effect of bad decisions
- Modeling with incomplete information - how do we handle this?
- Can we use a layered approach e.g participatory? ABM worked well for the electricity market simulations
- Problems are multi-objective, be wary of presenting extremes, or could use Pareto optimal thinking, or best guess of analyst. People are now more risk averse and need to think of robustness.
- Indicators we use are the ones we feel comfortable analysing
- Are models one-off or or more general?
- important decisions at the city level are made at the project level
- Water - too much supply side thinking - hard for governments to counter too - demand side would benefit from ABM
Opportunities for collaboration
- Methodology for CSS tool applicability
- Case studies of successes and failures
- Would need to be accessible outside CSIRO
- Matrix type presentation
- Liason with D_CITY (John Fraser at QUT)
- Visualisation
- Access to resources and expertise
- Relationship with universities
- Availability of students
- Data Sharing
- Forum on how to do it
- Data cleaningProtocols
- Interoperability
- Relationship with CIRAD
- Ontologies
- Modeling Urban Systems
- Hydrology
- Stocks and Flows
- Agent representations (a database ofr agent types?)
- Peri-urban land use
- Modeling urban change
- Urban components, interrelationships and implications of scale
- Model Development and Improvement
- Identify tolls as shared platforms
- Develop mapping for all components of urban systems
- Multiple' use tools
What is self-organisation and to what extent might it influence a city's development?
- Top down Design (e.g. planning) versus bottom up Self-Organisation (SO)
- Transport behaviour is self-organising - there's no controller on its collective behaviour
- Planners set constrants, boundary conditions on the SO possibilities
- Pattern formation dictated by constraints - tweaking the rules i.e establishing the right constraints
- Planners tend to concentrate on the hard system (physical issues)
- Self-organisation occurs within communities of people (the soft system) that make use of the hard system.
- Slaving principle of Haken (ref) - slower processes of change can constrain and "enslave" the faster processes but still leave a degree of creativity as to what can happen
- Positive feedbacks - growth inducing
- Negative feedbacks - growth inhibiting
- Can S-O suggest the amount and location of green space in a city with local versus global constraints?
- Systems of human settlements may be self-organising
How to start a holistic analysis: top down or bottom up?
- how holistic can you get? Answer: there may be no such thing as a completely holistic analysis only one which is more or less holistic.
- First need to identify the objective
- Holistic analysis by integration requires a bottom-up deeper understanding of systems
- Knowledge is usually filtered in some way. It is useful to have an interative approach whereby brainstorming is followed by critical evaluation of what's included or excluded and then repeat.
Aspects of the holistics system:
- Complex systems can exhibit behaviour that is commonly found in the resilience literature. for example: the greater the overhead of management or the more entrenched the tradition/ bureaucracy / rules, the more unexpected is the breakdown of the system (perhaps because of a belief that more rules engenders a greater security). When feedback loops are extended so far that we lose the signals that are telling us things are going wrong, then the false security from not receiving any negative feedback, actually drives the system towards an unexpected outcome.
- Crisis as a driver for change or the influence of leaders, key stakeholders, pioneers.
