This episode is a conversation with Roland Kupers about the concepts and vocabulary of complexity.
Episode notes
Roland Kupers - Bio
Roland is an independent advisor on Complexity, Resilience and Energy Transition.
He's Dutch and speaks 4 other languages fluently.
Originally a theoretical physicist he’s worked in business management in AT&T then in Shell where he held several senior positions.
He has written books on scenario planning, resilience and complexity in the context of public policy. And he’s writing another one on climate change at the moment.
He lectures in complexity and he’s an affiliate of Institute for Advanced Studies in Amsterdam.
Selected quotes
"And actually knowing what you can’t know is really, really important."
“…[in] the research offices of the central banks, you’ll see some junior scientists [were] working on this stuff before 2008 but, of course, nobody listened to them.”
“A very small intervention that leads to a massive change - and that, I think, is the prize for understanding these things better."
"We lead our private lives entirely under complexity, right? Nobody would think that the success of a party is the sum of the people that are there, the music and the booze, right? There’s something else. It’s the interconnection."
"The ... way to think about a complicated problem is to divide it into parts and solve the parts then put them together but that absolutely doesn’t work when you’ve got a complex situation."
More information
Transcript
Mark: Hello, I’m Mark Foden. Welcome to the Clock and the Cat, a podcast of conversations about clocks and cats obviously, but crucially, about complexity. The Clock and the Cat explores the emerging science of complexity, ultimately to help you generate better ideas and make better decisions whatever you’re involved with.
Mark: This is episode three and I’m going to be talking with Roland Kupers about what complexity actually is. But, before that, if you don’t know what the Clock and the Cat is about, please do go back and listen to episode one for a seven-minute, no messing, introduction to the podcast. If you did go away, welcome back. Here I am with Roland Kupers. Roland is an independent advisor on complexity resilience and energy transition. He’s Dutch and he speaks four other languages fluently, five if you count American. Originally a theoretical physicist, he’s worked in business management at AT&T, then in Shell where he’s held several senior management positions. He’s written books on scenario planning, resilience and complexity in the context of public policy. He has actually got another one on climate change brewing now. Roland and I have worked together on several projects, mostly to do with resilience. And in fact, we co-hatched the idea of the Clock and the Cat a couple of years ago. Roland’s an absolute wiz on complexity and I’m sure you’re going to enjoy the conversation.
Mark: So Roland, hello.
Roland: Hi Mark.
Mark: I’m in the UK but you’re in Singapore. What are you up to out there?
Roland: Yes, well I’m in Singapore to actually finish a book on complexity and climate policy that I’m wrestling with. And the other thing that’s interesting is Singapore’s is a real hotbed of complexity and I’m here, a fellow at the Complexity Institute at one of the universities. And it’s such an important topic for the country that they’re looking at actually massively increasing the funding for the institute because they really consider it strategic for the future of Singapore so that’s a really exciting time to be here.
Mark: Well it is exciting. I mean you did recommend that I went to the Complexity Institute early this year and I had a really good time there. It’s amazing how advanced they are and the Singapore government is in terms of complexity thinking. It’s a good place to be to start this conversation, I think.
Roland: Yeah, I mean one of the people who’s very close to the government here, tells me that every single government minister can answer the question, “What is complexity?”
Mark: That is cool.
Roland: Yes, and probably quite rare, I would say, around the globe.
Mark: Yeah, I’d say it was. Shall we kick off? First question, what actually is complexity Roland?
Roland: Yes, well I think that’s a question you can answer in many different levels. What, for me, it is, it’s the name of a development in western science over the past couple of decades. And the reason that development is important and super fundamental, I think is because it connects with an age-old idea that we celebrate in literature and in art and in music and in nature is that everything is interconnected with everything. Which is a wonderful image and a wonderful statement, but in our world, you actually want to give that a scientific reality. So I think that’s what, at it’s most fundamental level, is what the discipline of complexity is.
Mark: And you often hear the word complexity in association with complex system, or complex adaptive system. Where does this come from, this idea of the complex system and complex adaptive systems?
Roland: Well, what we refer to as a complex adaptive system is a number of parts that interact. They can be anything, whether it’s people and economy, birds in a bird flock, or all sorts of other elements, traffic cars in a traffic flow. There are independent elements and through their interaction, they have some sort of collective behaviour and they adapt off of each other. I think that’s what we refer to as a complex adaptive system.
Mark: So what’s the adaptive bit then, how is traffic and then let’s say in the city, how is that adaptive?
Roland: Well the intuitive rendering it is, is that the whole is more than the sum of the parts. So that what you have is you cannot understand traffic flow just from the different pieces. The famous example is these phantom traffic jams that occur. That, all of a sudden, without there being too many cars on the road, you can still have a traffic jam. And that’s because of the interaction and the way you get these kinds of spring effects and waves that go through traffic that cause these traffic jams. So it’s a collective behaviour that actually isn’t easily attributed to the individual elements of the system. It’s not magic. You can understand it, you can model it but it’s literally in a precise way that the whole is more than the sum of the parts.
Mark: Okay, so the adaptive bit might be that if these sorts of phantom traffic jams occur regularly in one particular spot, drivers might choose to take a different route through the city, is that right?
Roland: Yes, it occurs at multiple levels, there’s that, but in the traffic jam itself comes from because people adapt off of each other’s behaviours. Somebody hits the brake a little bit too hard, then the person behind them brakes in a slightly different way. So people adapt to each other and the result is that collective behaviours, the phantom traffic jam. It’s unlikely to occur at the same sport, the next day it might be somewhere else.
Mark: So the adaption happens at a sort of micro-level, or multiple scales or what?
Roland: Yeah, at micro-level and at multiple scales depending on the system, yes. But the essential thing is that we always assume that if something big happens, that there must be some big cause. And in these complex systems, the wonderful thing is that that’s not necessarily the case, you can have these collective properties that are not embedded in the individual bits.
Roland: The cliché, every discipline has its cliché right, and the cliché is bird flocks and complexity, but their cliché is not without a reason because they actually say something. So a bird flock, and everybody’s seen these kinds of wonderful murmurations of birds over a lake in the early morning light, and you see these amazing shapes that these birds exhibit. And yet we know two things about these birds, is there’s no choreographer and the agents in this system are the birds and they’re bird-brained, so they’re not particularly smart. And yet, collectively, they manage to do this extraordinary thing.
Roland: Now what complexity science then does is take a computer and try to program and simulate a bird flock and through that, try to explore what’s the minimum mechanism, what’s the root of this extraordinary collective behaviour? So you create a bunch of birds in a computer program and you have them adapt and react to each other in really simple ways and you see whether that reproduced the flocking behaviour. And it turns out that you can do that.
Mark: So that was one of the original simulations wasn’t it, what was it called? The Boids?
Roland: Yeah, Boids, yeah. It was possibly a scientist from Boston.
Mark: So you can imagine people in government and organisation struggling with complex problems. And I get a lot of people excited about the idea of being able to model and understand what’s going on better. So over the last 10 years or so, we’ve had a much greater computer power, we can actually do some of this. Can you just say something about the modelling that you can do on complex systems and why it’s different to sort of ordinary analytical modelling?
Roland: Yeah, I mean, one of the essential things is that it really helps you get a better insight into the dynamics of the system. But the one thing it probably won’t help you do is predict and that’ll be frustrating for many people because there’s a whole industry of modelling that’s aimed, of linear modelling, that aims at prediction. But the essential thing about complex systems is that generally, you can’t really predict how they will behave precisely. You can say something about classes of behaviour and how you might statistically influence their behaviour but prediction shouldn’t be your goal. And that’s the first threshold to cross for a policymaker and to somebody who promises to give them a prediction.
Mark: So if you can’t predict, what do you get out of the modelling?
Roland: You get a much deeper understanding of how the system works and what tweaks you might make in order to get to behave differently. A great example where actually… There are a couple of fields, not that many, where complex methodologies are really mainstream, one is dealing with epidemics. And it’s probably familiar to people from reading the newspaper if you talk about something awful like the Ebola epidemic et cetera. Is that you look at contagion rates, you look at containing, at reducing the speed of the spread of an epidemic.
Roland: And you can do all those things and you know that you will influence the outcome because you understand how the thing works intrinsically. What you’ll never be able to say is whether person X or person Y will become ill or not because that’s a statistical phenomenon. But you will be able to influence the behaviour of the whole.
Mark: So for example, if you go back to the birds thing, you couldn’t possibly predict the track of any individual bird but you do know that the group as a whole will exhibit that kind of swooping murmuring behaviour?
Roland: Yes, and you can exclude some things. You can demonstrate that they will never fly in a perfect cube for example.
Mark: Right.
Roland: And you will know why. So there are things that you can exclude. But indeed, knowing the exact behaviour of a particular bird or even exact shape that the flock will take when it takes off, you can’t know. And actually knowing what you can’t know is really, really important.
Mark: Yeah. So just to provide some mental hooks about this modelling, so what are the names of the methods and what are they about?
Roland: The main method is called agent-based modelling and there’s a whole mathematical apparatus that goes with that, but in concept, it’s really quite simple. Instead of a traditional model, if we talk about a traditional model where you write down a set of equations and you solve those equations and that will tell you how your system will behave. In this case, what you do is you take a computer and as with the birds, you create a program where you get the different agents, whether they’re cards or birds or people in the economy or anything else and you have them interact and you watch what happens. And then you tweak the rules and then you watch what happens.
Mark: It’s interesting, when I was in Singapore in March, there was a demo of the transport model, which is an agent-based model, and they had people, cars, trains all modelled. It was amazing to see this sort of visuals of this where a train would come into the station and a whole lot of people would rush out of the station and get on buses and in taxi’s and so on. And actually, they’ve used it to do modelling of problems of a train breaking down, that kind of thing. I just sort of think great, great example, at least visually, of this kind of technique. Have you seen that?
Roland: Yes, not this particular one but I’m aware. Yes, because crowd modelling is the movement of people whether it’s through a station or if there’s an emergency in a stadium or et cetera, those are perfect things that you can use these techniques for.
Roland: One thing to be aware of though, is not to oversell the modelling capability, right, because it’s, in some sense, fairly simple policy problems that you can model, so the movement of people or epidemics and those kinds of things. If we look at things like an economy, or climate change policy, or dealing with any quality or what’s the best healthcare system, those things are well beyond our modelling capability. But they are still complex systems and treating them as non-complex systems will certainly be wrong.
Mark: I think there’s a real drive to turn complex problems into simple ones so that you can go and solve them. So knowing the limitations of these methods seems to be really important. So good point. So that’s agent-based modelling and the agents being the birds, the cars, the people or whatever. And so we’re making a model of that, that shows the patterns of the behaviours as they interact.
Mark: There’s another method, another modelling method, called network analysis. Could you say just something about that?
Roland: Yeah, another way of generically representing complex systems is by representing them as a network and just like agent-based modelling, in many ways is quite simple. You map the interconnections between the different elements of a network and that unveils that structure of that network and how it drives collective behaviour.
Roland: A particularly nice example that I’ve come across is, I recently saw an article that people had taken the body of law of the European Union, which is about 50 thousand laws, and created a map of the network of those laws, so how they’re interconnected. So one law is interconnected with another if it refers to it, So that you imagine that you can draw this big spider diagram of which laws are interconnected with other laws.
Roland: And the fascinating thing is then when you apply some basic mathematical analysis to that network, you see that it has the same kind of properties as some natural systems have. So it’s not just a random set of connections, but it’s actually evolved in a remarkably organic way and it has particular properties.
Mark: So the network of the laws is a scale-free network?
Roland: Yes.
Mark: So what does that mean then?
Roland: A scale-free network is kind of the gold standard of networks. They’re the most efficient kind of networks that exist. And this is why nature likes them, because nature, in some sense, is either lazy or economic. It tries to organise things with the least possible effort, but the greatest possible complexity.
Roland: And so a scale-free network is one that has a bunch of big nodes and lots of less interconnected nodes. The internet is an example of a scale-free network. There are a couple of very large nodes and they’re connected to a bunch sub-nodes that are then connected to nodes that have much lower connectivity.
Mark: So some nodes are more important than others and you only discover this when you do the analysis or what?
Roland: Yes, and one of the places, for example, where you could see how incredibly costly it is to get it wrong was in the financial crisis in 2008. The World Central Banks were all worried about the health of the individual banks, whereas, of course, the thing that society cares about is the health of the banking network. You don’t really care if a single bank goes belly up, you really do care if the banking network fails.
Roland: After the financial crisis, there’s been a flurry of interest from central banks to understand the network patterns and the networks of interconnections between banks. Because what you’re after is a topology of interconnection that’s stable and more resilient to crisis then one that is not. And so it’s a completely different way of conceiving of a financial system is that you’re not thinking about, “Oh, as long as all my banks are healthy, it’ll be great.” But you’re actually worried about the health of the network as opposed to the health of the individual banks.
Mark: Well and that network including the, I don’t know, companies or individuals that actually have loans?
Roland: Well, in this case, there are various ways of conceiving that network. Basically, they look at… the base that people look at is the credit relationships between the banks. So who holds the biggest credit relationships with other banks because that’s typically where banking crisis unravel when banks have insufficient cover, so that’s the primary focus.
Mark: So this is interesting. This is a new application, when I say new, in the last 10 years, application of complexity thinking which I think is really exciting. But it’s not something that being part of the way of doing things before that, right?
Roland: Well, nothing is entirely… If you go back into even the research offices of the central banks, you’ll see some junior scientist who are working on this stuff before 2008, but, of course, nobody listened to them. And afterwards, they said, “Ah, we’ve got some people who understand this stuff.” And they were probably hauled into the board room to explain what it was all about. But I think that’s just the normal way the world works, right. It’s much more diverse and pluriform in thinking, but it’s dominated by particular approaches but it doesn’t mean the other approaches don’t exist at the same time. So it’s not that nobody had ever thought of it before 2008, but it certainly wasn’t the job description of the central bankers. And arguably, it still isn’t the job description of the central bankers, so we’ve not absorbed the lessons.
Mark: So one of the things we haven’t talked about so far, is this idea of emergence which is an important topic on complexity. So can you talk about emergence for a few minutes?
Roland: Yeah, it’s a really important idea. It describes this idea that the whole is more than the sum of the parts, right? That there are properties of a system that are just of that system as a whole. And the way it’s described is that those properties emerge from the interaction of the agents. And one particular example I came across recently, which I found really interesting, is somebody reframed the idea of decisions in organisations. You always talk about, “So who’s the decision maker and, which minister or CEO took, which decision?” But that assumes there’s kind of a top-down causality, somebody makes a decision and other people execute it.
Roland: If you think about a complex system, you can think of decisions actually emerging from an organisation through the interactions of all sorts of people. An organisation functions like an information processing system that comes up with a recommendation, with a decision. And then the role of the decision maker is a very different one, is that that’s the person who names and crystallises and formalises the decision. So decision making then becomes identifying the emergent opinion or tweaking it. But the fundamental process is that in the interaction of the organisation, a decision comes out, it comes about and then that’s named.
Roland: In a sense, the way a democracy works, right? People vote and then the result of the vote is identified by somebody and they say, “Well this is the winner.” And so it’s a very different way of looking at these systems and now focusing on decision making. But if the flocking behaviour of birds is called an emergent property of the underlying set of birds, for example. So it’s a really important concept also to get an intuitive stance of.
Mark: So this idea of decisions, I’m really excited about because I’ve got a sense that… I’m just thinking about an organisation I worked with in the past. There were probably dozens of times where I’d be walking down a corridor and bump into a few people having a discussion. Stop for a couple of minutes for a chat and suddenly find you’re in a 10 minute quite strategic decision. And then you suddenly realise that as a result of the conversation, you’ve understood something. And actually, a decision about doing something different is created there and then that had we not bumped into each other would’ve have happened. I think that kind of thing happens a lot.
Mark: As you say, a model is that we write an agenda for a meeting, decide all the things that need to be decided, go through the agenda and do it. But in most organisations, it simply doesn’t happen that way?
Roland: Yeah, I think there’s a difference between happening and describing, right? I think in most organisations, it actually happens bottom up in it’s an emergent process. But we just don’t describe it that way and that misrepresents the underlying process. I suspect, I don’t know for sure, but I actually suspect, if you take China, for example, most people refer to China as an authoritarian state where decisions are all made top down. And that’s why it’s so easy because you just got a couple of people with lots of power and they run the place.
Roland: I actually suspect it’s the same kind of misrepresentation. Yes, it’s a very centralised system. But I suspect that the decisions are pre-cooked in the same way our endless debates between cities and state-owned companies and individuals et cetera. And the central person does have the power to then bless a particular decision. But seeing a system like China as a much more decisions being much more emergent than top-down, I think is valuable and is a much richer framing of what happens in reality. And it also gives you a sense of the constraints. They can’t turn on a dime actually because they need to carry the system with them.
Mark: So the turning on a dime thing makes me think about the change in organisations and I know that lots of folks are listening to this podcast will be interested in change in their organisation. So it’d be really good to talk for a while about how that happens. And I’m thinking particularly this idea of tipping and face transitions.
Roland: Yeah, obviously the idea of tipping point got enormous press through this book a couple of years ago.
Mark: The Gladwell book?
Roland: Yes, The Gladwell book. But actually, I always think the more interesting thing is what happened before the tipping point. The tipping point is the result of something that happens before. Sometimes you compare it to a pot of boiling water, at some point the water boils. But actually, there’s all sorts of stuff that happens before so that suddenly it boils. And if you’re interested in change in an organisation, it’s the stuff that happens before that leads to a sudden transition that you’re interested in. Simply saying, “Oh, there was a sudden transition.” Is actually not very interesting. So what the study of complex system helps you see is what are the conditions, what are the things that happen before these sudden transitions. So first it helps you understand that sudden transitions are possible, which is a huge thing, right, because mostly we assume that everything’s gradual and incremental and linear et cetera. And having these discontinuities is not really in our vocabulary and is an uneasy thing. I think a great example is what you’re seeing now around the world, is these introductions of… the attempt to reduce plastic consumption, in particular, through one-use, throw away plastic bags. One of the early places where this happened was in Ireland, I think it was 2009 if I remember correctly. There was a deal between the Ministry of Finance and the Ministry of the Environment that they would introduce a very small tax on plastic bags and it would be great for revenue for the state. And the Environment Ministry would happy because it would gradually reduce plastic bag consumption. And to everybody’s surprise and slight sadness in the Finance Department, within three months plastic bag usage has virtually disappeared. So you had a sudden transition. Now wouldn’t it be nice if we could do those things purposefully. I mean it’s the dream of every policy maker to make a very small change that leads to a massive… a very small intervention that leads to a massive change. And that, I think, is the prize for understanding these things better, is how you can understand and how you can prepare and when is there is opportunity for these rapid changes.
Mark: So the plastic bag water was close to boiling and we just turn up the heat a bit, is that what happened?
Roland: I think it’s more a catalyst. The language is more from biology and chemistry. A catalyst is something you introduce in a system that creates it to change suddenly. I don’t know if you ever tried, if you put a bottle of beer in a freezer, it’ll remain liquid when you take it out. But if you knock it on the kitchen countertop, it’ll instantaneously freeze. And so this is what can happen in these systems is that a small pebble or small intervention or a small shock actually then changes the state of the system. So it’s not just adding a little bit of heat, it’s actually the thing that puts it over the edge in a sense.
Mark: Just to sort of talk about the catalyst event, my understanding of it, a chemical catalyst is something inert, a kind of platform for change if you like. Whereas I understand the thing that made the difference with the plastic bags was raising the charge. Raising the change for plastic bags, for example, in supermarkets, where you’re actually acting on the system, you’re pulling a lever rather than nudging some different behaviour.
Roland: That’s the question right because that’s what in our classical economic framing would lead you to believe, right. Is you, by changing the price, you change the utility function of people and then they’ve decided to change their behaviour. Except that it really doesn’t work that way. Because around the same time as in Ireland, they introduced this charge and completely eliminated plastic bags, they did the same thing in the Netherlands and everybody thought, “Oh, that’s cheap, let me buy another plastic bag.” And so people actually don’t care about the couple of cents that they spend on the plastic bag. But what had happened is that in Ireland, it went together with a phase, there was enough latent guilt or latent environmental consciousness that actually these plastic bags were a bad idea. And the small charge basically made the system gel to actually act upon that insight. But that had to be there before. And in a case like in the Netherlands, that wasn’t there and people just said, “Okay, well if it’s 10 cents for a plastic bag, I’ll just take two.”
Mark: So your complexity scientists will talk about starting conditions?
Roland: It’s the state of the network, right? This is where network theory comes in again, is if you understand the state of the network, you can change it. And there’s an initiative I think that was just launched in the UK where very recently they’ve agreed to characterise obesity now as a disease rather than as lack of control and eating too much. And in fact, there are some really interesting complexity studies on obesity that really show the network effects and really frame it as a complex system and how there’s contagion across the network and all these kinds of effects. And that allows you to then intervene in a different way.
Mark: It’s a book, I forget what it was called but I think it was Christakis and Fowler talking about how you’d be more likely to put on weight if your friends were overweight i.e. a network effect going on.
Roland: Yeah, there’s a network effect and there’s a real context effect also, social context also makes a huge difference. And this was a big thing, this is one of those naming things like decision making is that it was understood scientifically that it’s a contagious network issue. But we talked about it as if these were people lacking discipline and not being able to control their eating habits or not exercising enough. And actually now naming it as a contagious disease makes a huge difference because you can start using the policy tools that go along with that as opposed to just moralising or taxing or whatever you want to do.
Mark: So policy tools being something like network analysis, that kind of thing?
Roland: Yeah, and then just like when you’re dealing with a classic epidemic, you look at contagion rates and superspreaders and what are the behavioural change that will have a network effect as opposed to just admonishing people to go the gym.
Mark: So if you’re setting out on a big organisational change then it makes sense to try and at least start by thinking about what network effect might be in play.
Roland: Well, and first figuring out what your network is in the first place, right? The image that people draw of an organisation, the classic fork with the CEO at the top or their cute, they put the CEO at the bottom. But basically these linear relationships, that’s the way we talk about and we characterise an organisation. But that’s not what it is. I mean certainly, if you work in it, you know that that’s not the reality, it’s just a very small part of the reality. In fact, an organisation is a deep network, there are all sorts of cross connections. There are people who are not that senior who have a huge influence, who may be extroverts or maybe in some position of high connectivity that has a disproportionate effect compared to some very introverted senior manager somewhere. And so actually understanding the thing you have as opposed to the way you describe it is a really good starting point.
Mark: So I mean I’m particularly interested in this sort of organisational network and theoretically it’s fantastic to try and understand what’s going on in those networks. But getting the data for it is actually quite hard. For example, I knew someone who was doing some research on email data to see what emails was passed around in a big organisation, and actually, it’s really really hard to get that data.
Roland: Well my understanding is once you have access to the email data, that’s actually quite a powerful image. I mean you do some filtering on it to sort of get the bulletins and so on out and to different classes of emails. But there are some really classical studies about… and I think there are some standard tools now that allow you to map the network structure.
Roland: There’s an amazing paper that came out fairly recently where somebody looked at all the websites of the US Government, which are hundreds of thousands, and mapped the network of how they refer to each other. And this is all public information right? And it’s literally millions and millions of connections and it created a map and a network structure of the… and it’s a proxy obviously, right, and you have to worry whether the proxy’s a good one. But it did show which states and which cities were more interconnected than others and which organisations had influence across the country et cetera.
Roland: And so the data exists in company in the form of email and also in countries, in this example, in the form of websites. And if you can get access to that raw data, it tells quite a rich story. And again, it’s proxy of the real thing but at least you start asking questions about the real structure of the thing as opposed to the way everybody usually describes it.
Mark: And, of course, you’re not limited to modelling like things like, I don’t know, roles in an organisation or websites. You can connect people and organisations and laws and so on to actually get quite a rich picture of what’s going on.
Roland: Yeah, and the other thing that always never ceases to amaze me is that humans are actually very good at this stuff. That we’ve created a language and representation of systems that actually doesn’t leverage the almost biological talent that we have. We need our private lives entirely under complexity, right? Nobody would think that the success of a party is the sum of the people that are there, the music and the booze, right? There’s something else. It’s the interconnection. Whether something’s a success is if there’s a fabric that is an emergent property of a good party that’s created. And this is the way we deal with our lives and that’s what, as human’s, we’re really good at. So I think quite apart from the modelling and the science, et cetera, is being able to name and describe also our professional world in those terms, actually enables us to bring to the job the capabilities that we have and use them better.
Mark: And through an understanding of complexity, there’s enormous potential for improvement.
Roland: Yes, and debate, right? What I’ve always really found… I worked, as you described in the introduction, in industry for a long time and this is true in government as well, is its subject to these waves of fads. Somebody says, “Oh, everything has to be agile everything now has to be total quality or these things that wash over the system.” And this in a sense for me, this whole discipline of complexity is at least the beginning of a foundation underneath all of that so you can understand how those things work and why they work and even when they’re useful. Because these things are not useful in all circumstances, there are systems that are not complex and you shouldn’t use any of these tools and you should manage them very linearly.
Roland: So it also really helps you distinguish between the good and the bad and the fads and it creates a language where you can debate and have a little bit more precision about what you do.
Mark: So that’s I think one of the things that we try to do with The Clock and the Cat to get people to think about things that are complicated and things that are genuinely complex and to treat them differently.
Roland: Yeah.
Mark: I have these various ways of thinking about this, you know the other way to think about a complicated problem is to divide it into parts and solve the parts then put them together but that absolutely doesn’t work when you’ve got a complex situation. And knowing the difference and knowing when things are different on natural inclination particularly in organisations is to take an analytical reductionist approach. And how can we stop from doing that in complex situations?
Roland: Yeah, I mean your use of the word natural is kind of interesting because I would perhaps say we do that routinely but it’s actually, as I tried to say before, it’s actually not natural or natural inclinations to look at these complex systems. But we’ve been drilled. I mean we’ve all gone to school for 18 years or whatever, some big number of years and we’ve been educated in the method of reductionism over a good portion of those years. And reductionism is fantastic, I mean there’s nothing wrong with it. It’s just you need to apply it consciously and for the right things and not as a default approach.
Mark: We’re kind of running out of time now Roland. One of the things I did want to talk about was this idea of resilience. And you and I have worked on a few projects now where resiliency has been the focus. So how are resilience and complexity connected?
Roland: Yeah, resilience only exists in a complex system. And non-complex systems actually don’t have resilience properties and that’s because if you unpack resilience, you can unpack it in two things. One is robustness, which is just the ability to absorb a shock and not break. And the other thing is the ability to adapt and learn as a result of that shock. And that second thing, the adaptation and the learning is only something that complex systems can do.
Roland: Non-complex systems can be robust, just build a bigger wall and a stronger wall but that will never learn and adapt. And it’s an interesting debate, it’s all over the news these days with Trump’s obsession with the wall is a classic case, that immigration is a highly complex issue. And the last thing you want to create is a massive wall. You need a system that deals with the adaptability and the resilience of immigration. You need to engage with that, with that process and understand how it works.
Mark: The idea of resilience, if you have, again, thinking about this in organisational contexts. But if you’re in a complex situation, do you always want to be thinking in terms of resilience? Is that a good way of handling complexity? It seems to me that it’s a much more tangible idea. We really don’t want to have complexity, however, resilience sounds like a good thing to have. Is that a more palatable message in this day and age?
Roland: No, I don’t think so because what happens is people then confuse complex with the colloquial term of something that’s messy, you have a complex relationship with your brother-in-law or whatever. And that’s a fine meaning for the word, but in this case, complex with a big C actually has a different meaning. And it’s an unfortunate choice of word but here we are. And so if they can’t dissociate that emotion from the word, then we have a problem, and we do deserve to an extent. But resilience is just one property of complex systems. So yeah, if you’re interested in resilience, if that’s important for the particular organisation or system you’re concerned with, then that’s what you should look at. But if you’re interested in whether you have a scale-free network or not or you may be interested in past dependence, there are a number of these properties with complex systems that could be an area of focus. The other thing with resilience is to remember that it’s not something to maximise, right? You could have bad resilience. Organised crime is super resilient and you actually want to reduce it’s resilience so it’s not a property to optimise. In some cases, yes, in some cases, no. But again, it’s just one property of a complex system so I would counsel against saying, “Oh forget about this complexity stuff. Let’s just look at resilience.”
Mark: Okay, so that’s useful. So the one last question I’d like to ask, so the idea of the wicked problem has some currency at the moment. Is that exactly the same as complexity, a complex problem and a wicked problem are the same?
Roland: Yeah, I think so. And no word has exclusivity nor should it but I think complexity, for me, is kind of the umbrella terms over a whole load of things that have come before it. And it’s clearly not that one day people work up and said, “Oh right, complex systems.” It builds on systems thinking from the 1960s and Seller Automata all sorts of ideas and biology. People have written about wicked problems and even ideas like Agile and learning organisation, there are all these connected and associated concepts.
Roland: And it’s fine to use other words if they work better but I generally tend to think of complexity as the umbrella name that describes the entire field. And the reason for that is because it really is increasingly a scientific discipline with university institutes like the institute where I am now and most universities in the world now. At the Complexity Institute, there is a vast literature and you can go to people and help you with a particular problem. And there are no professors in wicked problems or very few, occasionally the university has but it’s not a discipline.
Mark: That’s interesting. So with that, if you’re interested in hearing more about learning about complexity, then Roland and I in episode four are going to have a conversation about learning about complexity so we can talk more about that then.
Mark: Roland, thank you very much indeed, that’s been hugely useful.
Roland: Yes Mark, always fun to chat.
Mark: So that’s it. Before we finish, if you found what you heard useful, please do subscribe. As I said, we’ve got Roland coming up in episode four again. Can I also give the important job of spreading the work because it may help someone else. And right now, before you forget, please message, email, tweet, slack or otherwise let your mates know about The Clock and the Cat. Thank you. Hope you listen again. Bye, bye.