The Leanpub Podcast 🎙 Feat. Yevgen Nebesov, Author of The Architecture of Power

Len: Hi, everyone. Welcome to this episode of the Leanpub Podcast, where I’m going to be interviewing author Yevgen Nebesov, based in Brussels. Yevgen is an independent systems architect, consultant, and speaker with more than ten years of experience across transport, logistics, and medical technology. His work focuses on sociotechnical complexity: how technical decisions, organizational structures, and human behavior shape one another. Yevgen is the author of the Leanpub book, The Architecture of Power: See What Shapes Outcomes, Find Where You Can Act. The book asks the question: why do unwanted outcomes keep returning even after people, policies, or priorities change? And we’re going to go into depth about the kind of theory and practice behind that. The book is split into two parts that roughly map to that. It’s a very fascinating book. In this interview, we’re going to begin by talking a little bit about Yevgen’s background and career, then we’ll talk about the book, and then we’ll talk about his experience as a speaker and writer and things like that. So, Yevgen, thank you so much for being on this episode of the Leanpub Podcast.

Yevgen: Yeah, highly, and thanks.

Len: I always like to start these interviews by asking people to talk about themselves and their origin story. So I was wondering if you could talk a little bit about where you grew up and how you found your way to Germany and into a career as a systems architect.

Yevgen: OK, yeah, it’s obviously a long way. I am originally from Kharkiv in Ukraine, where I grew up, and at the age of 18, I moved to Germany, where I have now lived for more than half of my life. I studied mathematics and informatics in Hannover. And after university, I started working as a software engineer, and over time I became lead engineer, then architect, lead architect, technical project lead, and executed roles like this. I spent the last fifteen years or so working in software projects, both in regulated environments such as medical devices and defense, as well as less regulated ones such as transportation, retail, and logistics.

Len: One thing that comes up often — a lot of Leanpub authors are people who move around a lot. I’ve moved around a lot in my life. And so I’d just like to ask: what was it like moving from Kharkiv to Germany at 18 years old? Was there a big culture shock? Was it seamless?

Yevgen: Oh, it was a lot of stress for me, because it was an age where you have your friends and so on, and then suddenly you lose everything and have to build it again from zero. But later on, I integrated into German society, and now I see myself as a German.

Len: OK. And did you speak German when you moved there?

Yevgen: No, I learned it here.

Len: OK, OK. And so you went from studying math at university to being a systems architect. What was that path like? Did you do more training, or did you just kind of learn on your own?

Yevgen: Actually, not at all. It was more or less a linear progression from software programmer to software designer, lead programmer, and architect. I read books and articles and watched talks all the time. There were no big jumps in my roles.

Len: OK, OK. And so, for people watching or listening, can you talk a little bit about what a systems architect is and does?

Yevgen: Basically, in software projects, you have — or you had previously — people who write code. Today, everyone writes code. But in principle, before you write code, you need to do some analytical work: how do we translate the business requirements into the structure of our application, how do we split the system into components, how do we define interfaces between them, which functionality goes in one component and which goes in another. This is the work that a software or systems architect does, in collaboration with the business, with stakeholders, and with people responsible for delivery. So it’s more or less a glue role — it glues all the other roles together.

Len: OK, OK. And so, is there — I’m just sort of curious about the working life of a systems architect. Does someone do the kind of notorious tap-you-on-the-shoulder, “do you have five minutes?” and then the next six months of your life is on a new project that came down from on high? Or, if you’re working at a big company, say, do you get to pick the projects you’re working on?

Yevgen: I usually work for one or two years on one project. I had sometimes two projects in parallel, but it’s really just the one.

Len: OK. And would a systems architect typically be there from the very beginning — like, some decision has been made to do something and you’re brought on board to design?

Yevgen: It depends on the company. A startup doesn’t have any systems architect — they just do things. But in a very regulated environment, for example in defense, you have systems architecture as one of the main streams, and you need to deliver some software architecture artifacts before you start coding. There are some quality gates that you cannot jump over. And that’s why in those projects, software architects or systems architects join even before the first developers start.

Len: Oh, that’s really fascinating, particularly in regulated environments. I know there can be all kinds of structures that exist.

Yevgen: It’s not fascinating — I know, I know.

Len: No, no, well, I mean, I know. But from a certain perspective, it’s kind of like — from the startup perspective, it’s just, “oh, do it.” Starting in something where there’s already structure that exists and that needs to be abided by along the way is actually an interesting challenge.

Yevgen: Yes. Especially because regulations can change as well, as we all know, along the way, or new regulations can be introduced.

Len: One question I wanted to ask you was — you know, one cannot do an interview these days without bringing up AI. When did AI — what we’re calling AI, just the loose way we use the term now — when did that first enter into your work? Was it a year and a half ago, like it was for startups? Was it five years ago?

Yevgen: I think, if we consider under AI the LLMs, like ChatGPT and upward — I started using ChatGPT, I forgot when it came, end of 2022, something like that.

Len: That sounds right, yeah.

Yevgen: And I was not a power user at the beginning — just translating some texts and writing emails. But over time I used it more and more, and today, of course, I use it almost the entire day.

Len: Yeah, it’s very — I would imagine it’s very useful. In addition to the sort of use we all kind of make of it, asking questions and stuff like that, I imagine — I’m just kind of interested from the perspective of someone who’s designing a system — the kind of non-deterministic feature of having LLMs involved in things. It’s funny because I’ve talked to some people who are like, “oh, you know, as a programmer, dealing with something non-deterministic is totally unfamiliar to me, and I had to learn that.” And other people are like, “yeah, people are non-deterministic, and so are lots of other things, and so fundamentally it’s not actually that new a thing with respect to being non-deterministic.”

Yevgen: One of the drivers for my book was exactly the fact that we now have AI with the capability of human-level intelligence, which means intelligence becomes a commodity — more or less, in certain disciplines more, in certain ones less. But for the last fifteen years, I was basically an intelligence trader. I exchanged my knowledge and experience for money, and many people — everyone in my industry — does the same. But now this intelligence becomes a commodity. And the question is, what is still human? And my answer to this question is agency — the ability to change things, not in software, but in people. And this is what is still human, I would say, and this was my reason to work on this project. Like, OK, we lost intelligence to AI, we lost dexterity and strength to robots — what is still ours? And that is agency. And my book is exactly about this.

Len: So, this might be jumping ahead too much, but it’s very interesting. The idea of, you know — because people will talk about having a human in the loop, things like that. I only know what I’m about to mention from podcasts and reading the news, but you know, military weaponry — keeping a human in the loop is something that’s actually very important to people. And it’s very interesting because there are conflicting views on that, right? Like there are people who are like, well, we’ve already had systems that can react incredibly quickly to circumstances for a long time — you know, once you drop an old-fashioned bomb, it’s autonomous. It might have some kind of gyroscope or something in it, or something more complicated. But one thing about having a human is that a human can actually be rewarded and punished in a way that machines kind of can’t be. Do you think that it’s the reward and punishment that is the remainder of human agency within systems? I’m bringing this up — this is my words, not yours.

Yevgen: Yeah, well, I think it’s connected — it motivates, it mediates human behavior, of course. But we are also carriers of trust. And this is, I think, what makes up our agency. For example, if I trust you and you recommend me some book, I will probably read it, compared to if some algorithm recommends me that book. And this is our ability, using our trust as a carrier, to change the behavior of people around us — not in a manipulative way, but to structurally change outcomes around us.

Maybe an example. Imagine we are in the year 2050 and we want to build a bridge. We don’t need an architect anymore because AI can design a bridge perfectly. We don’t need contractors anymore because we can source all the materials. We have self-driving cars to our construction site, and the robots can build the bridge for us. So we don’t need any people at all, people say. But do we have the bridge now? No, because there are still some stakeholders — people we have to convince to provide financing. We need to build a road that affects many buildings, and there are many stakeholders, and we have to bring them around to the idea and create a coalition of people supporting this cause. So there is a lot of political work to do. And this is a certain modality where we still need human agency to operate. The first one — the architect — was intelligence. The second one — the building — was dexterity and strength, which will be taken over by robots more and more. And the third one is human agency, where we operate not on the software, not on the hardware, but on people.

Len: OK, thank you very much for that great explanation and that great example. So with that stage set, let’s kind of go back a few steps and start talking about your book. And I think what you just said will make a lot more sense when we get to the end of the discussion — not that it didn’t make sense, but the technical terms that you were using. So the book is The Architecture of Power, and you start with this great question: why do unwanted outcomes keep returning after people, policies, or priorities change? And one thing that I love about that question is — when you talk about unwanted outcomes, you can mean in a personal relationship, it could mean a company, it could mean a software project, it could mean anything. It’s very deliberate, the level of generality at which the question is asked. So, why do unwanted outcomes keep returning even after people, policies, or priorities change?

Yevgen: Because in my opinion, outcomes are not a property of people or certain policies or rules. They are emergent properties of an entire arrangement consisting of people, consisting of rules, consisting of roles of people, beliefs, some rituals, public agendas. All of that constitutes some arrangement that reproduces outcomes. And of course, we can change a couple of nodes in this arrangement. But if those arrangements are stable, then they will for sure reproduce what was before. And my goal was to understand these arrangements — what are the nodes in them, what are the relations between them, and how they keep outcomes reproducing — and of course how to change those arrangements, even if you don’t have full control over them. If you can only reach a couple of nodes, how do you propagate your agency, your cause, throughout these arrangements?

Len: OK, OK, OK. So let’s start then by talking about the Power Loop. And by the way, there’s a great website accompanying all of this that walks you through things step by step, with practical examples and things you can do to test this out. So can you talk a little bit, to begin with, about the Power Loop?

Yevgen: Yes, the Power Loop is a central model of the book, and it explains power as a flow over six domains — I call them domains. The first domain is beliefs. It is how we see, perceive, and interpret the world — what things we keep as real, what we treat as knowledge, what we consider normal, what we think are things worth persisting and aiming for, and how we identify ourselves in this world. These are beliefs, the first domain.

The next domain is institutions. In the Power Loop, institutions are a triple of rules, roles, and rituals. Rules is basically any policies, any rules, any protocols that we have — it could be in an organization, it could be informal rules in a family, like who brings the kids to school. Then we have roles, like software engineer, or parent, or student. And we have rituals — for example, a daily standup, or a weekly routine, or even a routine in a family.

And beliefs and institutions affect each other. Beliefs enable the establishment of institutions and make them legible, and institutions, through their rituals, rules, and exposure, reproduce our beliefs. And of course, if there is a non-congruence between beliefs and institutions — if people don’t believe in the institutions — the institutions become brittle. They can survive on coercion for some time, but if there is no natural congruence between them, it’s just a matter of time until either the beliefs become congruent or the institution becomes disrupted.

Then institutions operate agendas. Agendas are arenas of public attention, of public salience. This is about which issues we are talking about, what concerns we want to solve. For example, a government is an institution and it has an agenda — every government session has a list of topics it discusses. This is the agenda, the third domain.

And on the agenda, topics are being filtered, they’re being framed, and only afterwards they reach the next domain, which is decisions. This is where we discuss points. And many people associate power with the ability to make decisions, but in fact, decisions are very boring, because what reaches the phase of decision was already filtered and framed at the agenda stage — and the agenda was operated by an institution that has its own structural bias, and the institution is based on beliefs. So basically, controlling a decision is just controlling the tip of an iceberg. There is much more below it.

But decisions create commitments that set actions in motion and flow into the next domain, which is events. Events are basically things that happen — or don’t happen — provided they were registered. For example, if it rained, that is an event. If it didn’t rain, then it’s an event if someone registered that it wasn’t raining. Or if a company fired fifty employees, that is an event. But if it fired fifty employees and no one noticed, then it is an occurrence but it is not a political event in the sense of the framework, because things that weren’t registered don’t count for political reasons.

An important thing in the domain of events is that they don’t interpret themselves — they are just raw material for the next domain, which is narratives. In the domain of narratives, we pick some events and connect them into stories. And we use these stories, according to this model, for two reasons: either to mediate our beliefs and shape them, or to bring points onto agendas.

For example, we observe that the climate has changed, we track events like daily temperatures — these are our events. And we connect them into the narrative: “Look, there is global warming, and it is causing a lot of suffering.” And we use this narrative to bring this issue onto the agenda. And then if it’s placed on the agenda and prioritized, there will be decisions — or no decisions. Again, new events, new narratives. These narratives can mediate our beliefs. Beliefs go into institutions, or narratives are used to mediate agendas. And this is how power flows throughout the loop and keeps some things sedimented in institutions and beliefs. And these sedimented things are basically then the arrangements that keep outcomes reproduced, if those arrangements are stable.

Len: That’s a fascinating system that you’ve presented.

Yevgen: I was very fast. I hope it was more or less understandable.

Len: It was. That was very clear. I’ve, for my sins, read a lot of philosophy, and so there were lots of things being set off in my brain there. But I really appreciate the structure of it. I guess one question, then — and this might be a hard or an easy question — and this is the name of your first chapter, which is: what is power? Within this structure, what is power? Is self-awareness a feature of power, for example? Or is this essentially a kind of force that drops out of an existing structure? Does power have a telos, an end, in this conception of it?

Yevgen: It’s a hard question. And in the book, I deliberately — even in the chapter “What Is Power” — I intentionally write that there will be no definition of power, because there are like a hundred definitions and all of them are bad.

Len: Yeah, no, that’s a smart answer.

Yevgen: Because they’re not universal. They define some aspect of the phenomenology of power but don’t define other aspects, which makes it a contested concept — it is one of those things that we all intuitively understand but cannot define. But the expression about power that I like is: power to society is what energy is to physics.

Len: Oh, sorry — can you say that again?

Yevgen: Power to society is what energy is to physics.

Len: OK.

Yevgen: So it is the driver of relations, of change, and change in the physics of society.

Len: OK. That’s a very good answer. I’m going to have to chew on that one for a while. One interesting thing that your description of the Power Loop made me think about was the concept of arbitrariness, particularly in politics. One of the origins of contemporary democracy came not accidentally after the Enlightenment, when reason became an important feature of the way people thought about how things should work. And one thing that, if you read Thomas Paine and other revolutionary leadership figures, they would complain about when it came to monarchy or aristocracy was arbitrariness — that the monarch could just decide something. Now, obviously, the monarch can’t decide whatever they want without a system around them, because it wouldn’t work without beliefs and loyalties and stuff like that. But what really bothered the kind of burgeoning notion of people who wanted democracy was that they wanted there to be sense to things. And that required at some level an explanation that made sense of why things were happening, why things were going the way they were. And I guess, in your discussion of power, is there a layer where it’s important for things to be explained and understood by the people within the system? Or does that not necessarily matter? Can someone just happily go about not knowing or caring why they’re doing what they’re doing, as long as they’re getting their paycheck or their satisfaction out of it?

Yevgen: So, first of all, the framework intentionally refrains from morality. It’s not immoral in the sense of being aimed to do bad things — it just takes morality out of the equation. In that sense, monarchy and democracy are just different forms of institutions, and the phenomenology of the Power Loop would probably be accurate in both cases.

But to your second question, I think the goal of changing the architecture from monarchy to democracy — the question is, do we need to explain it to the stakeholders? And my answer is yes, because this is where we step into the narratives, where we start cultivating the beliefs. Narratives impact beliefs, and beliefs are the precursor of new institutions. And we need narratives to hijack the agenda of old institutions with new points that disrupt institutions —

Len: Not in the negative sense.

Yevgen: I understand, yeah. We need to disrupt old institutions and bring new ones. And we need to do two things: we need to prepare people for the new institutions, and we need to destroy the moral superiority of the old ones by bringing new points onto the agenda.

Len: So essentially, a narrative is necessary to bring about a revolution through changed beliefs about the way institutions should work. Would that kind of make sense?

Yevgen: I think yes, yes. Of course, the question is always, what is the institution? In many cases, the framework doesn’t say an institution is something like a government. It could be something formal, but everything that repeats over and over again is, in my definition, an institution — if it has some recognizable rules and rituals, then it’s an institution. It could be a family. And the question is always, what is the scope of the things that we want to change through our revolution, or evolution, or whatever? And of course, if we call this thing an institution, then according to the framework, yes, we need to feed new narratives into the beliefs of people in order to build this incongruence between the beliefs and the institution.

Len: OK, OK. So, with a little bit of a theoretical framework now — the Power Loop — can you talk about power mapping, now that we’re armed with these ideas? What do we next do with them?

Yevgen: Yeah, basically the end goal is to understand how to change power arrangements around us — for example, how to replace monarchy with democracy. And before we start, we need to map the current situation. I was thinking a lot about how we map power. And because there was already the Power Loop with its six domains, and every domain has its own substructures — like five types of beliefs, three pillars of institutions (roles, rules, and rituals), and four dimensions of agendas — so every domain is split into subdomains. Why not use that structure to map the power arrangement, by going through every domain and subdomain and asking the question, OK, what’s happening there? What are our rules? What are our roles? What are our rituals? What are our beliefs of each of the five types? What is getting admitted onto the agenda? What is getting priority on the agenda?

The idea was to take the Power Loop — the six domains, which depict a dynamic flow — and make a static snapshot of the current state of the loop by bringing them onto the canvas with six fields, one for each domain, and answering questions about every substructure of the Power Loop and seeing what is happening there right now. So it’s always a snapshot, because power is always in flow.

Len: And so to make this kind of concrete — one example you give on your website, which I’m looking at right now, is luxury as a power landscape. I don’t mean to put you too much on the hook for something, I don’t know how long ago you wrote it. But it’s interesting — I brought up government and revolution, but one of the examples you give is luxury handbags. How do you make an extraordinary price intelligible? And I was wondering if you could talk a little bit about that. I mean, from a certain perspective, a cheap handbag and an expensive handbag, both in a sort of everyday “carrying things around” sense, perform the same function. But an everyday handbag and a luxury handbag are obviously very different things to people who conceive of them differently.

Just one little anecdote before I ask you to go through the example. A friend of mine was working on a doctorate a few years ago in political philosophy, working on Isaiah Berlin. His supervisor was friends with the Berlin family, I think. And as a gift, he received Isaiah Berlin’s book bag — one of Isaiah Berlin’s book bags. And he came to my birthday party at a bar in London with this book bag, put it down beside him. It was a very busy, popular bar, and about ten minutes later, he looked down and his book bag was gone. Someone had stolen it, probably hoping there was a laptop in it. My friend took it like a philosopher. But the interesting thing about that experience was that it went from being Isaiah Berlin’s book bag to just being a book bag — the person who stole it probably looked inside, didn’t find a laptop, and threw it in a bin. Anyone who found it might have thought, “Oh, neat book bag,” or something like that. But I wanted to tell that story because your example of the luxury book bag and what makes it a luxury book bag brought that story to mind.

Yevgen: Yeah, it’s actually the same example, which highlights that the value of the bag is not in the bag itself, but in what people think about this bag. In your case, for your friend, this book bag was special because Isaiah Berlin used it, and it was of high value. For someone else, it was just an empty bag. And the question in the world of luxury is: how do we create the entire arrangement of power that makes people believe that a certain object is loaded with value? Which means that when I buy a Gucci bag — if I would buy a Gucci bag — I’m not buying the bag. I’m buying the effort that Gucci invested in making everyone around me believe that this bag has value. So it preloads this object with value by cultivating the belief that this is something. And of course, if I go to the Himalayas, that bag has no value at all.

My goal was to understand how this entire arrangement of believing in the value of this bag is being reproduced — what are the beliefs behind it, what are the narratives used, what are the topics on the agenda, what are the decisions usually being made, and what are the events. And I went through the entire canvas, through all six domains, and mapped what is happening there.

Len: Yes, and you talk specifically about what you call landscape mapping. I’m looking at it right now — decisions, events, agendas, narratives, institutions, and beliefs — the six dimensions you were talking about before. And it actually is — just for people who are listening or watching — you do map these things out, as it were, on paper or on a screen, in the canvas, and you fill it out. And one thing that I found very engaging when I was going through the book and the website is that the most important thing, I think, to begin with, is to define what the outcome or outcomes are for you. Is that right? Before you can do this kind of power analysis, you have to decide what matters to you in the analysis you’re doing.

Yevgen: Exactly. For example, you work on a project. In my particular case, I’m working as a systems architect, and my concern is the quality of the things we create, because quality is crucial for the long-term success of the project. And, for example, I don’t like the outcomes being produced in the project — that the project always decides to be quick and dirty rather than putting in effort and doing qualitative work. So this is the outcome I want to change. I define the things I don’t like, I build a canvas explaining how this has been reproduced, and then I define my goal outcome — we do qualitative work — and then think about how to move the current landscape in small steps toward the landscape that will produce the outcomes I want.

Len: One perhaps difficult question I’d have about this is: how explicit — and I guess it would depend on the circumstance — but how explicit should one be, in a company, say, about one’s actual outcomes? Because a kind of classic problem with changing systems from within is you may have an enemy, someone who wants you to fail, just because they don’t like you. Or there could be a competitive environment — two people at the same level, only one position above them, and if both want to get there, they might try to prevent the other person from getting there. So I’m just curious about whether concealment — when one is doing landscape mapping, one might not want to let everyone know what you’re doing every day. Do you know what I mean?

Yevgen: Yeah, yeah, I see. I think that’s already very tactical. The framework is more about structural changes, and the question of how to jump from one state to another depends on the current situation. In certain situations, of course, we need to keep silence and just move silently through the night, just to prevent political friction from happening. And in other cases, we need to explicitly state the goal to mobilize a coalition on our side. It depends on the particular situation.

But what I explicitly differentiate in the book in Part Two — I think there is already a chapter on it being published — is the difference between structural tactics and activation tactics. The structural tactics are about what exactly do we want to change — for example, we want to build something, we want to change something in the institution, in the rules, in the agendas, some structural change in one of the six domains. And another dimension is activation — it is about which participation from our people, from other people, we need to achieve for a given structural change.

For example, in some cases I want just people to start using the frames that I’ve introduced. For example, the project doesn’t pay much attention to quality, and I start using the term “technical debt.” And because it’s “debt,” it has some emotional connotation. What I need from other people in the first stage is just to use this term — that’s one form of participation. So one dimension is how do I activate people, and another dimension is what do I activate them for. And every structural tactic has different forms of activation it needs. For example, if I want to introduce a new ritual — which is part of an institution — I need other people to act, they need to join this ritual. Or if I want to introduce a new term, then the needed form of activation would be for them to adopt it.

So, back to your question about concealing or explicitly stating — explicitly stating would be a form of activation to mobilize other people for your structural move.

Len: Yes, OK, yeah, that makes a lot of sense. And now we’re moving on to the second part of the book where you set out the tactics — activation tactics, event tactics. The one that jumped out at me is narrative tactics. That sounds like we’re surrounded by narrative tactics in our lives. Can you talk a little bit about what narrative tactics are in the context of the lead architect project that you talk about?

Yevgen: OK, so in the Power Loop, narratives are about creating stories that can help move agendas — bring some points onto agendas — or changing beliefs. And they do so by selecting certain events and putting them in causal form. There are of course different tactics we can use. We can highlight certain events, we can select a subset of events that help our point, and so on. These are all different tactics.

And of course, if I want to achieve in a project that we pay more attention to quality, I cannot start with narratives. I always need some events, because events are the fuel — they’re the data for the narratives. So the first thing I need to do is start tracing problems that are created by the old outcomes — starting a log of delays, of cases of the system being unavailable due to technical problems, and so on. I need a body of evidence that I can use later. And then in the narrative domain, I would pick some things from this body of evidence and construct a story that will help me advance my cause by linking these events to some outcome or structure that I don’t want anymore. For example: “These five system crashes happened because two years ago we made this decision,” and I want to use this evidence to reverse that decision. That’s the mechanics behind this tactic.

Len: And you give some structure to the narrative tactics side, which is quite fascinating. The narrative jobs of name, explain, assign, project, and anchor. I wanted to go into a little bit of that — you can go into as much detail as you’d like.

Yevgen: As I said, I introduced in the domain of beliefs five types of beliefs. And these five verbs in the narratives actually map to those five types of beliefs. These five types of narrative jobs are designed to change or operate in one of these belief types.

We have ontological beliefs — it is about what things exist. For example, you’re from Canada, right?

Len: Yeah.

Yevgen: I would say Canada doesn’t exist in reality — it only exists because eight billion people think it exists. If, for example, all of them lost their memory for one day and woke up tomorrow, what is Canada? Never heard of it, right? So these are our ontological beliefs about what exists. And there are lots of conflicts about naming things — the most recent ones are about what something is called, just about giving things names. And this is the category of narratives that are designed to affect exactly this type of belief — just to name things. Like, there is no such thing as burnout — we constructed this concept. Or “toxic masculinity” — our DNA doesn’t know this concept, it’s purely constructed. Or Germany — it’s a human invention, it doesn’t exist. All these things are narratives that someone constructed.

Len: Childhood is an example I always like to give when I’m trying to explain that.

Yevgen: Yeah, exactly. And the next type of beliefs — epistemic beliefs — it is about what is true, which sources we consider as true, which methods of gaining evidence we consider as satisfying to establish what is true and what is not. So it is about — the first is about which things exist, and when we have these atoms on which we can construct knowledge, we go to the epistemic beliefs, where we say this is true, like Canada exists or Canada is large. Right, I cannot argue about Canada if Canada doesn’t exist.

Len: Yeah, yeah.

Yevgen: So “name” maps to ontology, “explain” maps to epistemology. And then we have social order beliefs — it is about what is expected behavior, what is normal and what is not normal, what people respect, what is bad, and so on. They basically regulate us as a society. And there are narratives designed to mediate these beliefs — like saying “this is good,” “this is bad.” All these value-laden beliefs.

Len: You know my book better than me.

Yevgen: Well, I’ve got it right in front of me.

Len: But yeah, so name — ontology; explain — epistemology; assign — society…

Yevgen: Yeah, and then the next one is project. That maps to the next type of beliefs — teleological beliefs — about what goals we believe we want to achieve, what sacrifices we accept to pay, and so on. So basically when the narrative “projects,” it says, “Our goal is to deliver this project by the end of the year.” This is setting the teleological beliefs through the narrative.

And finally, maybe the most important type of beliefs — identity beliefs. It is about who we are, who they are, who I am, and so on. And there are narratives that cultivate such beliefs. For example, “We are a team that is responsible for good quality,” or “We are a nation that always keeps our streets clean.” Basically our understanding about us and about them, whatever they are. And there are of course many narratives today that address those beliefs — they anchor our identity.

Len: Yes, yes. So the term for that narrative tactic is “anchoring,” and I now understand that maps onto the belief about identity, which makes a lot of sense. Those are very well-chosen words, I think, to map to each other.

So I guess you’ve kind of already made it clear that on a certain level, you don’t believe that these things are kind of real — for example, in the way that a lot of people would just take it for granted that, of course, Canada exists. But you would actually claim that Canada doesnoesn’t exist, in a certain sense.

Yevgen: Depends on the definition of exist and not exist, yeah. I would say it’s subjective. And I think there is maybe no single person who would say Canada doesn’t exist, which makes Canada universal. But there are countries in this world that are not universally recognized.

Len: Right.

Yevgen: And this is where we can argue about whether the name of a country exists or not. But obviously there can be other kinds of — I mean, any concept can be challenged on these grounds, right? Any concept with a name can be challenged.

Len: Like gender, for example, can be challenged on these grounds. Like, well, what do you mean by that? And as soon as you can ask “what do you mean by that,” something is being separated from being merely in a natural, universal state, right? As soon as you can ask “what do you mean by that,” this complication is just inevitably there in the way we think about what exists and what doesn’t.

Yevgen: Exactly. And the question for me is not what exists and what does not exist. The question is, why do we need this concept at all in our language? And my answer is, well, we need this concept maybe somewhere to be used in institutions — to define roles, rules, or rituals. For example, if you take the example of gender and sex, the question is, why do we need sex? Because, at least, there are some medical treatments that are just for men and just for women, and this is the reason we need sex. And we had rules like there is a men’s toilet or a women’s toilet, and this rule reproduces our need for this concept of sex. Or there are language words like “he” or “she” that differentiate — so we create some downstream legacy concepts that keep the dependency on the initial concept in place and keep it reproducing.

And then we say, well, sex is one thing, but we also have gender. And we say gender is equally important as sex. And then we say, let’s now introduce institutions based not on sex but on gender, because it has equal standing — it doesn’t matter if you have a physical chromosome or not, if you perceive yourself as a woman or a man, it has equal standing to having an X or Y chromosome. So we started defining rules based on gender. And what happens next is, well, what if I identify today as a man, tomorrow as a woman, and the day after tomorrow as a coffee pot — which makes me something unstable — and this renders all these rules as senseless, because an institution is something that needs to be stable. And I think this is a moment where we have to question whether this institution makes sense at all. And if this institution operates on some invented concepts — and both sex and gender are invented — we can actually ask ourselves, do we need those concepts? If we’ve just proven that this institution doesn’t make sense and these are the only usages of these concepts, do we need these concepts at all? And then we can maybe delete the concepts of gender and sex entirely if we don’t use them in any institution.

Len: Yeah, this is obviously raising tons of very fascinating questions. You reminded me — I was once in a seminar with the philosopher Richard Rorty, and it was very interesting. He got into this kind of argument with a young graduate student who was in physics, I think. And what we were talking about was making a statement like “Canada exists,” and what I recall Richard Rorty saying was it makes no sense to also say “it is true that Canada exists” — there’s no need to say “it is true that Canada exists,” you can just say “Canada exists.” And I remember this scientist, he was just infuriated by this. No, it makes a difference to say “it is true that Canada exists” as opposed to just saying “Canada exists.”

Yevgen: Because in terms of the framework, one would be the ontological belief and the other one the epistemological.

Len: Exactly, exactly. Yeah, and for him there was clearly a clear border between the two of them.

Yevgen: Yeah.

Len: Yeah, and it’s also interesting that on a certain level — and I think I’m bringing this up for a good reason — there is a thinking person that is inevitably a part of this framework. And I think from the scientist’s point of view, the important part of saying “it is true” is that I am committing myself to this in a sense, but also I’m saying that it’s true independently of me. That was what was so important to him — to say “it is true that Canada exists” means that it’s true independently of me. But that’s very different from the framework where if we all forgot about it — or on another level, if all conscious beings in existence ceased to exist — I think most people would not object to the claim that Canada would no longer exist.

Anyway, I just bring that up because it raises the question that in the framework you’re giving, there has to be, to some extent, a conscious thinking being engaging with it in some way — a carrier of the belief.

Yevgen: Yeah, I think — I’m not a philosopher, I’ve never studied philosophy — but I think there is this concept of modernism and postmodernism. Modernism was about splitting things that are really real — what you said, that exist independently of our beliefs — like my phone exists. And there are purely constructed things, like Germany. And they were trying to place the world into real and constructed things. And the movement that came after, postmodernism, was like, “OK guys, everything is constructed, so let’s not even try.”

Len: Yeah. There’s — maybe I’m wrong —

Yevgen: No, no, I think you captured the spirit of it.

Len: I mean, not to be too specific about it — postmodernism began as an architectural movement — but you captured the spirit of the thing. And like, this is kind of a little bit of an idiosyncrasy of mine, but the real key flip there was existentialism. And so if you read Sartre and Camus and stuff like that — going back to Kierkegaard and stuff like that — the idea that was really fundamentally introduced is that it’s inescapably a matter of personal will and action and responsibility, everything that we engage with. And so from the existentialist point of view, there’s this concept called bad faith, which relates to things that I’m responsible for — and the term “responsible” is doing a lot of work there — but relating to things that I’m responsible for as though I’m not responsible for them. And so when you see people clutching to the idea that something really exists that is in fact inevitably some kind of construct, that’s bad faith, right? Like, “No, no, this is — you know, it’s Adam and Eve, not Adam and Steve,” or whatever. No, you know, it really is this way, there really is one way that things are and it exists independently of our action and will and thought and belief. That’s where that split comes from — and again, this is my idiosyncratic way of looking at it — that’s really where it comes from. And that’s the great human divide in the way we approach the world, right? And that’s one of the reasons I think I probably intuitively brought up revolution and the switch from monarchy to democracy. It’s like what happened to the democrats was they were like, “That’s just a man. He can give whatever order he wants, that king, and it doesn’t mean anything if we don’t choose to follow.” And then there are other people who’d be like, “No, that’s the king. What are you talking about? You can’t just declare that the king isn’t the king. Who are you to do that?” And of course that’s what the existentialist collapses down to — like, inevitably it’s you. You’re the one person that you are, you’re engaging with the world, and you can just choose not to relate to that king as a king at all. Power structures might force you one way or another, but inside, internally, you can just relate to anything however you want. And that’s kind of when a lot of people get very disturbed — what they might call postmodernism, their sense of disturbance about that, is exactly what the existentialist would zero in on as bad faith and fleeing, a form of retreat from the responsibility that you actually have. So basically it’s a form of aggressive cowardice.

Yevgen: OK. But basically, back to your example with the king — I think it’s very well aligned with the ideas of the book, that power is not a property of a person, it’s not a possession of a person or of an actant, it is a property of an arrangement. The king’s power emerged because there are certain rules and beliefs around him, and everyone — even a slave — is connected to this arrangement. And I think it goes to your argument that even though we cannot control some things, we have some things in our immediate arrangement, some nodes that we can reach, and this makes us responsible — not maybe for the entire outcome, but for part of it.

Len: Yes. Yeah, yeah, definitely. I mean, the kind of extreme version of that is Camus and The Myth of Sisyphus, essentially addressing the question that since you can kill yourself at every moment, and every moment that you’re alive you’re responsible for — and in a certain level, even with respect to the things that happen to you, because you could choose to end it all — so even with respect to the things in relation to which you are passive, you are playing a role, and you have a kind of agency with respect to those things, because you could just check out.

So, in the interest of time — this is a fascinating discussion, and I know we could talk a lot longer. And by the way, the book is just so good at explaining these things, and you can tell from the conversation that there’s a system behind it. Once you kind of trust, “OK, this is going somewhere,” it’s a very powerful book. But in the interest of time, and in the interest of our audience, moving on to the last part of the interview, where we talk about the author’s experience as a writer — I was wondering, and I mean this in a very concrete way, how did you write the book? Did you block off time in your calendar? Are you someone who, when you’re traveling, goes to a café? Do you go into a study or an office? Because I think when people are starting out and thinking, “I want to write a book,” they’re like, “But how do I do it?”

Yevgen: Yeah, I think there are many things that affect my book writing. But first of all, I consider writing itself like writing code — because before you write code, there are maybe two years of work, of research, and just writing the code is the least problem. And I spent a lot of time before I started writing the book by researching and iterating the concepts. And after the Power Loop emerged as a stable concept, I started working on the book. Before the Power Loop, I had one or two years of just similar constructs — my goal from the beginning was to explain power in just one model, and it took me dozens of iterations to get there.

But once the Power Loop emerged, I started writing — without a real structure at first. I basically used writing to verify the ideas, because when you write and you see it doesn’t fit, then you go back and start thinking again. And there were a lot of cases where I was already at sixty or seventy percent of being finished, and then I reverted back to the first and second chapter because I needed to rethink everything again.

But right now, I don’t write a book at all, if I can say that. I try to build a knowledge system according to — I don’t know how familiar you are with this — a programming pattern we had twenty years ago called Model-View-Controller, which separates your idea from the representation. For example, we have an Excel table with some numbers — that’s the model. And then we have a pie chart or some other chart — those are views. And we have a controller that builds the view from the model.

I follow the same approach. The book is just one of the views. My website is another view. My LinkedIn posts are another view. If I write articles, just another view. And the model is a set of concepts. And I mostly work on these concepts. Over time, I created a lot of translation skills that translate those concepts to multiple views. For example, today, if I work on a tactic — right now I’m working on the tactics for agendas — I create it in the model, and then I have very well-trained skills that translate it into the website representation in terms of content, structure, and layout, and also into the book. And then I just review the views, the outcomes, and feed the result back into the model.

I started using this approach already after the model was stabilized, because earlier I really needed to write the book linearly to test the concepts. Now when the framework is set, I just bring new notes and translate them into different views. And I can do it — my problem is that I have work, I have three non-profit organizations, I have three kids, I never have time. But I have like ten minutes here, ten minutes there, and ten minutes somewhere else. I needed a system that would work in such an environment, without four hours of blocked focus time — I don’t have that. And by this approach, by splitting the model from the view, I can change some small things in the model without needing much time for it. And then let things translate into the views, and in another ten minutes I would review one of those views. And this helps me survive and succeed with the book.

Len: That’s a fascinating story of a book coming to existence, and then what it became. I just love the idea of it organically growing into a structure, but once it’s a structure it’s something different than it is when it’s the growing organism. And I love also the idea of facing the fact that it has to fit into your life the way you’re living it. The creative project that you’re engaged in — I think that’s actually a very deep insight that a lot of us don’t have when we’re first conceiving how we’re going to do things. That it must — whether that means you need to change your life or something like that — a project needs to live within the framework that you’re living yourself.

Yevgen: It was a necessity. I would love to sit eight hours a day and just write a book, but I don’t have that possibility. Well — not yet, but let’s see.

Len: So, again, thank you so much for taking the time to do this interview with me today for the Leanpub Podcast. The book is The Architecture of Power: See What Shapes Outcomes, Find Where You Can Act. And I’m just so glad to learn that this is the middle, I guess, of a wider project. The book is fascinating. The way the description you gave of how the narrative tactics map onto the five belief dimensions was just amazing. And the rest of the book I know is full of things like that to find. So I highly recommend it to anyone who’s interested in how companies really work, how projects work, and how you can affect change within them. So thank you very much, Yevgen, for being on this episode of the Leanpub Podcast.

Yevgen: Thanks, Len, for having me. And it was an insightful discussion for me as well, because you brought so much interesting stuff that I didn’t know before. Thanks.