Transcription
[Music]
Kinevin is a sense-making framework. It's not a model, and that's actually quite an important distinction. A model seeks to represent reality. A framework is a way of looking at reality. So, Kinevin is a framework. It has five domains, all right? It's drawn with a series of curved lines. It's fairly easy to do. And it also picks up on the original three types of system: ordered systems, complex, and chaotic systems. But what it does is divide order into two and adds a fifth domain of disorder.
Now, I'll start with this order, then run around the others. So, disorder. At the center of the framework is the state of not knowing what type of system you're in. When we first create Kinevin within an organizational context, this is often the biggest domain because people aren't used to an ontologically diverse approach to sense-making. They're used to single ontology, not multi-ontology. So, basically, disorder is the state of not knowing. Are you ordered? Are you complex? Or are you chaotic? Dependent on your own personal preference, your own personal history, you'll tend to assess the situation based on your own preference for action. Bureaucrats, for example, generally assess problems as failures of process or assess the need for new process. Experts generally assume they didn't have enough time to do the investigation. Politicians are pretty good at complex. They get lots of people from different backgrounds and look at different possibilities. And of course, the fascists love a crisis because then they can be given absolute power to tell everybody else what to do. There's a reality on the cognitive science front that we assess a situation based on how we've already decided to act, which is why an awful lot of Kinevin techniques are about preventing you from having any knowledge of action when you do situational assessment or any opportunity to divide it. Either way, disorder is an unhappy place. It's a state of being ontologically ignorant. It's a state of not knowing what type of system you're in. Therefore, you're likely to get things very wrong. It's not the same thing as a chaotic system. In a chaotic system, there is no possibility of order. In a disordered system, it may be ordered, you just don't realize it. Right? So, that's the criticality of disorder.
Order divides into two: um, obvious and complicated. Now, in order systems, there is a linear relationship between cause and effect. The same thing will happen again, the same way, twice, and not by accident, but by the nature of the system. The difference between obvious and complicated is that in obvious, everybody can see what the relationship is and nobody disputes it. So, in the United Kingdom, for example, we drive on the left, left-hand side of the road. In the United States, they drive on the right-hand side of the road. There are historical reasons why this happened, but nobody really disputes it anymore. You don't drive on the right in the UK or on the left in the US. You accept it. It's in the UK, you know, the UK is in the category of countries that drive on the left. So you change your behavior accordingly. So, the decision model is: sense, categorize, respond. Yeah, which type of, which country am I in? You know, okay, it's the UK. That falls into the category of countries which drive on the left. Right, I'll act. Drive on the left. Nice and simple. It's standard operating procedures, doctrine in the army. It's a good space. Over-constrain it, you get the collapse into chaos because actually you're creating a constraint which is impossible to pursue. The metaphor, a bit, if you said you must never ever drive on the right-hand side of the road in the UK, and that means if a child runs on the throat in front of you, you should kill them rather than avoid them, and that's obviously a nonsense. All right? So, even the most rigid systems allow for exceptions, or at least they should if they're properly constructed.
In the complicated domain, however, here it's not self-evident, other than possibly to experts. So, if you have defects or teas in the field as an advisor for you, it may be obvious, but for the decision-maker, it's complicated. They know there is a solution. They know you probably know it, but you still have to convince them of your expertise. Now, that may be in a high-trust situation, just telling them. It may be doing analysis. It may be doing investigation. The decision model in a complicated domain is to sense, analyze, respond. So, I gather income in data, I analyze it, that tells me what to do, or tells me who to call in to tell me what to do. And whereas in an obvious domain, I can apply best practice. There's one right way of doing things. In a complicated domain, I apply good practice. There are different variations, dependent a little bit on context. And a major mistake that organizations make is to actually impose best practice in a good practice domain. Doctors, for example, know that there are variations which in exceptional circumstances they would take. Them to only adopt one approach is a major mistake. You need some degree of variation in the system.
Now, there is actually an interesting related problem here. Experts can be trusted within the bounds of their expertise if the situation is complicated. But in a complex domain, they only create hypotheses. And the danger is a lot of hypotheses are wrong. Nothing wrong with that, it's the nature of the system. The trouble is, if we believe a complex system is complicated, and the expert gets it wrong, we lose trust in expertise. So, the boundary between complicated and complex is important because on one side of the boundary, somebody with the right training, the right qualification, can be trusted. On the other side of the boundary, they can't. Again, take a medical situation. In single morbidity situations, one expert may know everything. In multi-morbidity situations, you've got all sorts of interactive elements. It may be more complex. It may actually require a different approach, a multi-team approach, for example.
So, ordered, ordered systems: obvious and complicated. Sense, categorize, respond, or sense, analyze, respond. Apply best practice, or apply good practice. And coming back to our theme of constraints, in an obvious domain, I have rigid or fixed constraints. In a complicated domain, I have governing constraints. I, the boundaries are defined, but within the boundaries, within the containment, a greater degree of variation is allowed.
In the complex world, well, here it's multi-modular. There's no linear relationship between cause and effect. I don't know what the right solution is until I act. So, the decision model is: probe, sense, respond. If I think the situation is complex, and remember the heuristic, the evidence supports conflicting hypotheses that can't be resolved within an acceptable time scale, then I test or search out ideas which are coherent and ideas for action which are coherent. And I run parallel, safety-fair experiments. And the experiments reveal what's possible. In a complex world, everything I do changes the situation. So, a diagnostic is an intervention, and an interventionist diagnostic. What generally happens is the experiments merge, mutate, and change. It's less that one succeeds or one fails, but in combination, they change the space so that things are easier to manage. So, the complex world: probe, sense, respond. And the practice is emergent, or it's ex-adaptive. Something we're going to deal with later in the course. Another way of describing acceptation is radical repurposing. Um, so, for example, I have found many things I can repurpose to open a beer bottle in a hotel bar late at night when I haven't got a proper opener, yeah? Repurposing is something human beings are good at. In biology, that's called acceptation. So, in a complex world, we generally discover novelty through repurposing existing capability. So, we don't apply best practice, a good practice. Practice is emergent, it's ex-adaptive.
In a chaotic domain, on the other hand, there is no pattern. There are no constraints. There are zero constraints. Everything is all over the place. So, my first action is to create a constraint. Or if I want to move things into the complex, into the chaotic domain, I get rid of the constraints. And the irony is, that may require creating a container. It's kind of like, like a sandpit, all right? A scope works project. I hide this thing away and I remove all of the constraints, but I protect it. We'll talk about distributed decision support in a different module. But in a chaotic system, there are no constraints, which means I will get complete novelty. But I act. I don't analyze. I don't probe. I don't categorize. I act. Sense, respond.
Now, the boundary between order and chaos is shown as a cliff in Kinevin. The reason for that is, if I over-constrain a system which can't naturally be constrained, I move to that boundary zone, also known as a zone of complacency, and then I have this catastrophic failure. And the trouble is, when I fall over the cliff, it takes very little energy to fall off a cliff, but it takes a lot of energy to climb back up it. So, it's something you should try and avoid.
A final distinction here is between enabling constraints and governing constraints. This is the complex-complicated divide. So, say a governing constraint contains what we do. It's like a series of rules or principles, yeah? Or less principles, more rules. An enabling constraint is something like a heuristic, a rule of thumb. Now, this approach revolutionized warfare at the time of Napoleon. He created a very simple heuristic: march to the sound of the guns. Until Napoleon, people basically waited for commands which came from the general who sat on a hill on a horse and sent out runners, or individual commanders took autonomous action, hoped they got away with it. Napoleon created the heuristic, and the heuristic in this context is measurable: marched to the sound of the guns. That meant there was a guiding principle you had to follow, but you knew other people would follow the same principle. So, the system had self-organizing capability. With the US Marines, they have three: capture the high ground, stay in touch, keep moving. Now, these are called enabling constraints. And the way you manage a complex world is by discovering the enabling constraints which naturally exist within people's day-to-day narratives, codify them, attach them to teaching stories. And as long as people follow the constraint, the heuristic, then they're allowed to make mistakes, because mistakes are inevitable if a situation is complex. Now, again, we're going to come back to that in a separate module, which will look at heuristic management and the boundary between heuristics and rules. But for the moment, it's a key distinction within Kinevin.
Fixed constraints, governing constraints, enabling constraints, no constraints. Best practice, good practice, emergent or exemptive practice, novel practice. Dependent on the situation you're in, you make decisions in a different way. You assess the situation in a different way. You handle contingency and risk in a radically different way. Kinevin's huge originality is it doesn't seem to replace what went before. It ceased to create boundaries around it and create space for novelty on the other side of those boundaries.
[Music]