Lankelly Chase’s approach to working with complexity – with comments

Lankelly approach to change (pdf):


I’ve responded to this at a little bit of length because there’s no comments field on the Lankelly Chase website, and because:

  1. I care about this and think it is very improtant
  2. We are doing a very related piece of work on local systems change for save the children (which I hope will be published one day), so these issues are very live for me
  3. I’m about to go on a retreat to discuss local systems change
  4. I see a recurrence of some loose language and maybe some lazy thinking which I think is hugely risky
  5. I think just a little more grit in the oyster and realism about this would add a lot



Lankelly Chase’s approach to working with complexity

[with my comments in square brackets]

What does effectiveness look like when working in situations of complexity? How do we understand what to do or where to start? How do we attribute results to our actions?  These questions go to the heart of our feelings of competence and agency.

When you choose to contend with complexity, as we do at Lankelly Chase, you start to realise how professionalised structures help to keep such troubling questions at bay. Projects, programmes, hierarchies, funding cycles, milestones, service level agreements, budgets, key performance indicators all feed our need for certainty and order. The messy unboundaried web of interdependencies recedes behind a dense framework of process and structure, and with it our feelings of anxiety and helplessness.

Except, of course, the complex reality remains undiminished, and containing it with our structures proves as slippery as clambering onto an unruly lilo in a choppy sea. Worse still, our need for orderly structure can make a bad situation worse. It obscures our appreciation of the issues we need to address, leading us to act as if the situation weren’t in fact complex, and so denying the lived experience of people caught up in the web.

Lankelly Chase’s mission is to change the systems that perpetuate interlocking disadvantages such as homelessness, mental ill health and substance misuse. Until recently, our default mode was to fund interventions and projects for 3-6 years. At some level, we knew that life cycle problems could not be solved by grant cycle solutions, but as a relatively small grant funder what else could we do? Answering this question required us to face the possibility that despite our mission and good intentions we might be playing our own part in a perpetuating system.

For us, this came down to a judgement about whose needs were being met: were we funding projects, interventions and organisations because this was genuinely the most effective contribution a funder of our modest size could make or because we felt comfortable funding projects, interventions and organisations? While realism about our size has always been critical to how we understand our effectiveness, it became increasingly clear that it had led us to reduce our understanding of the problem to the scale of our preferred response to it.

As an independent funder, we felt we had a responsibility to resist reductivism: “if we can’t then who can?” But feeling responsible and knowing how to act are not the same.  We are so conditioned to operate within reductive structures that stripping them away leaves us with very little. The paralysing confusion and anxiety all come crashing in.

We have just published a revised description of our approach to change which sets out how we seek to take action in a situation of high complexity. The thinking behind it can be summed up as follows: Lankelly Chase has a growing conviction that the outcomes we seek can only happen through the actions of whole systems. Although there are many parts of a system – projects, workers, organisations, rules, funding, communities, institutions – that have a bearing on a particular disadvantage or harm, they are all continually affecting each other. No individual part exists or has an effect in isolation of the others. This leads us to think that sustainable change depends on the way all the parts interact.

[This means that we have to think about changes in the structure of systems – and the system behaviours and change approach have one particular slant on that. There is, of course, a distinction between reductive structures (I agree that these proliferate), and those which absorb complexity, or usefully attenuate variety]

What we find in the field of severe and multiple disadvantage is that the overwhelming majority of energy, attention and resource is dedicated to improving the parts of the system. This is understandable because those parts tend to be knowable and to an extent controllable, whereas the relationships between them take us into much less solid or definable territory. However, a collective focus on the parts has ultimately, in our view, been self-defeating because each can only be as good as its relationship with the rest.

Lankelly Chase decided to shift from this majority position to focus on growing the health of the relationships between the parts. Given how nebulous such a challenge could become, our first step was to ground our approach in the work of our many partners. We set about observing closely what effective relationships in situations of complexity actually look like, and what our partners needed in order to connect their part of the system effectively with the rest.  We looked at what they were doing in many different contexts, and consulted widely with many different system actors, until we were able to distil our learning into nine system behaviours. (Some people call these system conditions or even pre-conditions.)  We asserted that it was the presence of these behaviours in a system, rather than the action of any one organisation or project, that explained its effectiveness.

[I’m sceptical of some of this – see below]

So then what? Given that we were moving into largely unoccupied territory, we knew that we couldn’t map out outcome-determined programmes, not least because they are so ill-suited to situations of high complexity. We decided that our role was to create the infrastructure that would enable us and our partners to explore how to create healthy systems. We had a specific purpose in mind: to work out what it would take to embody the system behaviours at a scale that would ultimately make a difference to outcomes.

[We have to be careful with words like infrastructure, ‘actions of whole systems’ (when we’re talking about ‘the system’ being ‘unboundaried’), and embodying behaviours]

The methodology we agreed on was action inquiry, which we characterise as acting our way into a new way of thinking through continual iteration and learning. We now use our resources, along with the independence, flexibility and longevity they allow, to create collaborative spaces for our partners to explore ways of building the health of systems, increasingly at the level of a place.

[Note that independence, flexibility, and longevity are specifically divorced/outside of the context of the systems being discussed – this has some interesting (and useful) implications]

Frankly we don’t know where this new approach might take us, what ‘scale’ might look like or indeed whether the system behaviours will ultimately turn out to be right or helpful. Our current dream is of a critical mass of people thinking and acting in a more systemic way and thereby ‘tipping’ the way the whole system behaves. But we’ve no clear plan for how this might show up in the world, not least because we know best-laid plans don’t work with complexity.  So returning to the opening question, what would the effectiveness of such a strategy look like?

Our Chair Myron Rogers has coined a very astute maxim: “the process you use to get to the future is the future you get”.

[Interesting – I remember having a lot of discussions with marxists/socialist about whether the character of the revolution predetermined the character of the revolutionary state! And Conway’s Law – “organizations which design systems … are constrained to produce designs which are copies of the communication structures of these organizations —M. Conway”  – see discussion at . This is something it’s very good to take seriously, I think – but it can lead to the fallacy (which must have a good name somewhere!) of believing that however we design ourselves will generate the design of the system we’re designing. Unfortunately, there are some laws of nature/systems laws too which have to be taken into account – and utopian models don’t always translate into utopian outcomes. Worth looking at some of these other laws in that context eg.

A complex system that works is invariably found to have evolved from a simple system that worked. A complex system designed from scratch never works and cannot be patched up to make it work. You have to start over with a working simple system. – John Gall (1975, p.71) 

“The weaker the average regulatory ability and the larger the average uncertainty of available regulators, the more requisite hierarchy is needed in the organization of regulation and control for the same result of regulation… . [In other words], the lack of regulatory ability can be compensated to a certain extent by greater hierarchy in organization.”

And most importantly, the Conant-Ashby theorem (every good regulator must have a model of the system it regulates – based on Ashby’s law that only variety can absorb variety)

(and I put a whole bunch together here – ]

This feels right to us, and we have taken it to mean that we at Lankelly Chase also have a responsibility to embody the nature of the change we want to see. Once you start to see yourself as part of – rather than apart from – a system, then a lack of congruence between how you act and how you want others to act shows up pretty quickly. It turns out that if you concern yourself with the health of a system, acting effectively and acting ethically start to look pretty much the same.

In publishing an updated account of our approach, our aim is to expose our thinking to many alternative perspectives so that they can help reveal our blind spots, connect our approach with many interrelated parts of the system, question its boundaries, and above all challenge the privileged space we hold in an unequal system. All of these are ethical concerns that connect directly with how we can judge our effectiveness in a complex system. Organisations are rarely the best judges of their own ethics and effectiveness, and can persuade themselves of many things. So we genuinely welcome the inconvenient truths as much as the odd bit of encouragement.

Lankelly Chase Systems Behaviours

We have identified core behaviours that help systems function better for people facing severe and multiple disadvantage. Through observing different fields including homelessness, violence, health, the arts, community development, substance misuse and youth work, we’ve seen that (i) it is the presence of these behaviours, more than any specific methodology, that seems to account for positive change and (ii) these behaviours need to be present and continually promoted in every part of the system.

These behaviours are about perspective, power and participation.


  1. People view themselves as part of an interconnected whole

Everyone working towards positive change understands that their actions form part of a web of activity made up of the contribution of many others. Everyone wants the system as a whole to work, and knows they cannot control it.

  1. People are viewed as resourceful and bringing strengths

Everyone is viewed as bringing both strengths and weaknesses as part of a resourceful network of people who are continually growing and learning from each other.

  1. People share a vision

People appreciate each other’s perspectives and seek common purpose and understanding.

[I agree that these are Good Things, my only quibbles are: 

1- we should be careful to note that, at the very least, there are some negative self-fulfilling prophecies that quite powerfully ‘control’ systems – financial incentives and narrow targets control systems very ‘effectively’ indeed. And good systems do not need everyone to hold the beliefs indicated (though they are nice and possibly helpful to have) – people can work brilliantly in their silos and still be part of a superb system.

2- this is very powerful. But staringt from this, and from action inquiry (and from noting that perspectives really demonstrate whole experiences and shape behaviours) would be far more appreciative

3- sharing a vision is not the same as appreciating each other’s perspectives – the norm here is a ‘nice’ one so how does this engage with the real world?

The risk is here, that in privileging a certain type of systems view, we seek to norm that and judge or impose those who are not capable or do not want to or have other incentives.]


  1. Power is shared, and equality of voice actively promoted

All people are able to play their fullest role in building an effective system. Unequal distribution of power, including structural inequality, is continually addressed.

  1. Decision-making is devolved

Those people closest to a complex situation are free to engage with its uniqueness and context and to use their initiative to respond to it.

  1. Accountability is mutual

System improvements are driven by accountability to the people being served. The people being served are supported to take responsibility for their own change.

[This gets really challenging. These are norms, and they may or may not be appropriate. Sharing of power and devolution of decision-making and mutual accountability could lead to the worst of the breakdown of the hippy communes or death cults. And how does this norm work with those who actually have power? I would like to see these seriously inquired into and rethought. Equality of voice is not appropriate in all circumstances (equality of voice of abusers? In what context?). Devolved decision-making is ambivalent – surely this means appropriate devolution of decision-making? And what are the mechanisms that might bring about mutual accountability? Each of these assertions are value-less but are being applied with an implicit ethical lense, and that’s a dangerous fudge at times]


  1. Open, trusting relationships enable effective dialogue

People feel safe to ask the difficult questions, voice disagreement and deal with the conflict and uncomfortable emotions that surface.

  1. Leadership is collaborative and promoted at every level

Leadership is identified and valued as much in the person experiencing interlocking disadvantages and the frontline worker, as in the CEO or commissioner.

  1. Feedback and collective learning drive adaptation

People can see a learning loop between the actions they take and their understanding of the problem they are trying to solve, so that each is being continually adapted and refined.

[All good and good aspirations – and conflict and emotions are recognised in (7). What does leadership mean in (8)?]


We hold five assumptions about the nature of systems which come from our experience of supporting change across the UK:

  1. Systems are complex and often messy webs that are constantly shifting. They consist of tangible things like people and organisations, connected by intangible things like history, worldviews, context and culture. [Yes]
  2. Everyone who is part of a system holds a different perspective on its nature, purpose and boundaries. No one person holds the whole truth (including us). [There’s a risk here of sliding into thinking ‘and if we simply hold this true and bring all perspectives together, we can see everything]
  3. Everything and everyone exists in relationships, and these involve emotions. [Bloody ace]
  4. Change emerges from the way the whole system behaves not from the actions of any one project or organisation. We therefore need to help build the fitness of the system to generate positive change. [This is of course, at minimal recursive and potential self-contradictory. One project or organisation therefore needs to find ways to change the behaviour of the ‘whole system’]
  5. The complexity of systems means we can’t fully plan how to achieve the changes we seek, but we can identify several conditions that enable positive change and the actions that are likely to move us toward our goal.