This is a transcript of episode 399 of the Troubleshooting Agile podcast with Jeffrey Fredrick and Douglas Squirrel.

Have you got Systems Thinking all wrong, like the author of an article we read that got Jeffrey all riled up?

Listen to the episode on SoundCloud or Apple Podcasts.

Someone is Wrong on the Internet: One Good Part

Listen to this section at 00:14

Squirrel: Welcome back to Troubleshooting Agile. Hi there, Jeffrey.

Jeffrey: Hi, Squirrel.

Squirrel: So, as usual, I sent you an article, and you read it. And as unusual you’ve had a completely bifurcated response to it. And so I think the best thing is just to let our listeners hear this dual response. Let’s maybe start with the part that made you happy, because there’s another part that made you angry.

Jeffrey: Yeah.

Squirrel: And I’m very interested in what the article is about and why you had these reactions to it. And I’ll try to temper it a bit. How’s that?

Jeffrey: That sounds good. Now, the funny part was you sent me this article from, and I hate to even say the person’s name because it made me so angry, but the person is Ed Bradon.

Squirrel: Don’t hold back, Jeffrey! You know, really tell us what you think!

Jeffrey: And he wrote an article in an online journal called Works in Progress. And the article is called Magical Systems Thinking. And unusually, for the things that you send me, I’d already read this, because I read Works in Progress. I had come across this, and I’m a fan of systems thinking, and this is an article about systems thinking. And the title Magical Systems Thinking, that’s a good title, right? If you’re familiar with magical thinking, then I definitely want to hear about magical systems thinking. And then I read it and I was angry. I was very angry.

Jeffrey: But before I get into being angry, I’ll just point out there is one good part of the article, which is it introduces Gall’s Law, which says any complex working system evolved or came from a simple system that worked. And the idea is that you can’t go in directly build a complex system. Any attempt to do that will fail. And that’s such a great concept. I recommend that however people want to learn about this, do not read it from this article. Instead, they link to one of the other places on the internet. I think we’ll put in something from a site called Driverless Crocodile in the show notes–

Squirrel: Excellent–

Jeffrey: –Which doesn’t have anything much more than that. Yeah, but I like the name driverless crocodile, but unlike this article–

Squirrel: Hold on, hold on, hold on. Jeffrey. Jeffrey, let’s slow down for a second. Let me just make sure I’ve understood what Gall’s Law says. So I’m thinking of examples like the water system that delivers water to my house, or Facebook. And in either case, what Gall’s Law says is someone did not sit down and design all the pipes that go to all the houses in my town and figure out where they would all go, and then connect them all up to a single reservoir. Somebody started by running some water to one location, and then to another, and then to another, and then changed it a bit. Developers like to call this refactoring, although I think they’re using the wrong term, and gradually evolved it until it delivered water to lots of houses.

Squirrel: And the same thing with Facebook, which started providing a service to universities. And it started in Harvard, and then it went to Yale, and then it went to somewhere else, and then it went to a bunch of other universities. And now we all look at cat pictures on it.

Jeffrey: Yeah.

Squirrel: Because Zuckerberg did not plan out that he would take over the world in the way he has. He said, ‘I’m going to make university students able to meet each other in a better way. Have I understood Gall’s Law correctly or am I confused?

Jeffrey: Yeah, you have it right. And any complex system, like the internet itself, would have a very similar story. It connected a few universities, a few research centers, and then grew out. Arpanet became the internet. Or the phone system, or any number of other power delivery systems. They didn’t start as the gigantic, continent spreading or globe spanning system we have. Those things evolved out of simpler systems that worked, rather than being designed from the beginning as a complex system.

Jeffrey: And so what that might mean in software is you start with an MVP or you start with working elements. The good part about this article, and, you know, I hate to say it, is the idea that systems kick back. Systems have behaviors of their own. But that’s now getting to the part that I’m going to be angry about. So I don’t know if you want to…

Squirrel: Oh no, no, it’s fine, go ahead, be angry. I give you permission. We like emotions here on troubleshooting agile. Go ahead.

Jeffrey: Okay, good! Because the fundamental direction of this article, and this is where I get angry, is it’s denigrating the idea of systems thinking. If I want to be generous to the author, and I will be very, very briefly. On X, in describing this, he was saying– I can understand where he’s coming from. He expresses the concern that, say, the UK civil service guidance will say like, ‘systems thinking: anyone can do it.’ And his concern is that people will take some ideas from system thinking and sprinkle it on their very complex projects and now think that they can maybe make, high speed rail, High Speed 2, work the first time in the UK or the California high speed rail. That they’ll be able to make the system work by sprinkling on systems thinking.

Squirrel: For the listeners who don’t know, those are two high speed rail projects that are famously, massively behind and not at all benefiting from systems thinking or apparently any form of thinking.

Someone is Wrong on the Internet: What is Systems Thinking Anyway?

Listen to this section at 05:44

Jeffrey: Yes, exactly. Exactly. That’s his complaint. That would be valid. Some people will take simple mechanistic thinking and scale it up to the level of a system and think that their linear extrapolations will work. That’s simply not true. And this is why I’m angry: that is the fundamental insight and motivation behind systems thinking. So the article and why I hate it is because he’s saying like, ‘hey, that system thinking stuff is dumb and here are some failures of it.’ And his definition of system thinking is completely and utterly wrong! The entire point that he’s making in his article is something that you would get from a systems thinking book in the introduction, and likely off the dust cover, right?

Jeffrey: If he had ever picked up a systems thinking book and read the back of it, he probably would have had the thesis for his article. But it’s only his ignorance of systems thinking that allows him to go and write this article. Or alternatively, he is aware that he’s being a charlatan and misrepresenting it, but he thinks that he that it’s okay to misrepresent it, given that he has a good intent in mind. Either out of ignorance or poor decision-making he’s giving people a misrepresentation of what systems thinking is, which is to say, to understand that systems have emergent properties and that these emergent properties are likely to be ones that you don’t expect, and that therefore, it’s worth considering the properties of the system and the potential to have unexpected consequences. That’s why you lead by being cautious when you have complex systems.

Jeffrey: So as someone who values systems thinking and the idea behind it and the motivations behind it, to have someone so misrepresent it is really infuriating. Even if the directions and things that he points to, of Gall’s Law, and the fact that systems do have emergent properties that will, as he put it, the systems will fight back. That’s very true. Now, it would have been a great thing to say, but unfortunately he felt a need to misrepresent the entire domain as a way of doing it, for a reason that I don’t understand. I mean, I can speculate. So.

Squirrel: Okay. Okay. Jeffrey. Jeffrey. Jeffrey.

Jeffrey: That’s a bit of my rant.

Squirrel: Yes!

Jeffrey: Yeah. Okay, fine.

Squirrel: So we’ve we’ve heard Jeffrey in full flow. Let’s slow down a minute to let our listeners catch up. So the thesis of the article appears to be this notion that systems kick back. And Bradon gives some examples, and I’ll just quote one of them, which our listeners are probably pretty familiar with. They’ve probably heard of somebody named Obama, and they’ve probably heard of something called Obamacare, where America tried to change its health care system and set up a website called healthcare.gov. When the America tried to do that, it didn’t work very well and the website didn’t work and people didn’t get health insurance. And the thing was generally a failure, which we’re used to. Right? We see a lot of those. And you cited the high speed rail as yet another in a long, long, long list of systems that someone tried to design. Not like the examples I gave before when we were talking about Gall’s law. As if Mark Zuckerberg sat down and said, ‘I want to get cat pictures to a billion people on the planet. How am I going to do that? I will design an entire system from the start here in my Harvard dorm room.’.

Jeffrey: Yeah.

Squirrel: It sounds like that thesis is one that you agree with, but it’s actually a thesis that’s fundamental to systems thinking. Did I hear that right?

Read One of Those Instead

Listen to this section at 09:34

Jeffrey: Yeah. That’s correct. And as we’re talking about it here, I recently read a different book. I’ll mention two things. I was literally reading Thinking in Systems on the plane yesterday and thinking about this article and the book. Thinking in Systems is an introductory book on systems thinking, it was published in 2008 from a private draft that was circulated in 2001. So it’s, you know, pretty old. All to say, it’s not even intended to be like, cutting edge. It’s sort of ‘these are foundational thinking system principles.’ And there’s another book I recently read called How Big Things Get Done. And it’s particularly talks about the failure and pitfalls of large projects and what leads to them and what you can do instead. Don’t read this article. Go and read one of those instead. I think How Big Things Get Done would maybe be like the most practical thing. And it brings in particular ideas around estimation. You don’t estimate it based on linear extrapolations, but this is a class of what? What’s my reference class for estimates. So some really good thoughts about how to manage complexity in that book. But yeah, anyway, sorry… You were saying? What was your question again?

Squirrel: My question just was, have I understood it right, that the actual argument that Bradon makes, we agree with, and I certainly agree with, Gall’s law is right. If you build systems, try to build them from your dorm room to take over the world, they won’t work.

Jeffrey: Yes! And it’s the reason systems thinking exists as a discipline! The argument he makes is correct, and is the reason systems thinking is a term of art at all, because people experience the failures of trying to do this. Because it seems reasonable that I can just sort of scale up the kind of planning I’ve done in the past, and then that led to a bunch of failures, and people were like, ‘what’s going on here?’ And it’s like, ‘oh, it turns out these complex systems–’ and again, relevance to our audience here, like, what are we often doing? And our contexts will vary, but probably our audience is dealing with people.

Jeffrey: People by themselves are complex. And if you have a group of people, then there’s even more complexity, the complexity of their interactions. And then we’re building systems. These are often distributed systems. And we might have distributed teams of people building distributed systems. So we have complexity and complexity. We have all these systemic interactions that we need to understand. In fact, some of the things that we most foundationally talked about or the reason that we started talking about this stuff from the beginning was around the elements of culture and how conversation impacts that. And we were really looking at what’s the systemic effects of the kind of conversations we have on the relationships we have, and then the output of our teams and what we’re able to accomplish together. That’s why we started with communication, because it had such huge systemic effects.

Jeffrey: …Um, sorry, I got a little bit off-topic again, but yes! So there are points he’s making that are right, and that’s why systems thinking exists. And that’s a good thing to study. So the insight that he has in his article is correct. And the reason this stuff exists and should be understood. And then you read this person, you’re like ‘yeah?’ And then why do you like, try to claim that this is a novel insight and somehow different from the whole foundation? It’s sort of like someone saying, ‘you know, the problem with astrophysics is that they’re not looking enough into space, right? We don’t just live on this Earth, but the Earth is like in a whole universe. And those astrophysicists should think about the whole universe, not about Earth.’ You’re like, ‘wait a minute… What? That’s what the domain is about.’

Squirrel: Okay, let’s come back to Earth. [Background dog noise] And my dog agrees. So let’s see if we can come back, because there are great things that people can get from this article, I think. And so Jeffrey is saying don’t read it. I actually think listeners should read it, but with a sort of protective mask on, realizing that Bradon is mischaracterizing the very target that he’s talking about. But when he looks into the stars, he’s seeing some interesting stuff. And I just want to call out one particular one that I thought was really fascinating, and I wonder what Jeffrey thinks about it. That is a book by Gall which is called Systemantics, meaning antics of systems. And it has a whole bunch of very interesting little aphorisms, like ‘systems tend to grow, and as they grow, they encroach.’ And ‘the system does not do what it says it is doing.’ ‘The name is emphatically not the thing.’ So that sounds like a really interesting book, and yet another one that we should probably link to. So I think the message here really is, you should learn all about systems thinking. You might want to read Bradon’s article, with great caution, and it’ll point you to some very interesting examples of system thinking, which he’s attacking but actually agreeing with, have I heard all that right?

Jeffrey: Well, I mean, there are some things I agree with. I think that Systemantics book sounds like a lot of fun. I recommend that Squirrel, you and I put a link to that in the show notes along with the driverless crocodile, and we go in and direct our readers. Don’t read this article. Go read those other things instead. But this article, to the extent that it has original ideas: they’re bad, and the good ideas are not original and are better read elsewhere. That’s my take on it.

Squirrel: All right. There we go. I think we’re going to have Jeffrey lie down. He’s going to have to recover from his ranting on this topic. I’m the one who puts the links in. So I’m going to link to this article, and listeners can make their own decisions.

Jeffrey: Agh! Fine!

Squirrel: Haha, it comes with a health warning. You can read it or not as you wish, but I don’t think we can bash Bradon for 15 minutes without letting people actually form their own opinions because–

Jeffrey: Okay, fine.

Squirrel: –As you know everyone, we like it when people disagree with us! Even if they cause us to rant and put Jeffrey’s blood pressure up, and he has to put a cold cloth on his forehead. So if you’re Ed Bradon, we’d love to hear from you. If you are someone who thinks he’s right, we would like to hear from you. Please tell me I’m right, Jeffrey. We would like to hear from people who disagree with us.

Jeffrey: Absolutely I would love to have Ed Bradon on–

Squirrel: Yes.

Jeffrey: –And I would love to have him explain his motivation in characterizing the way he did. And I will say this up front, I can imagine that there are good reasons for him to do this, and I think it’s a little bit– Well, here’s the best analogy I have. If you’ve ever been someone, like myself, who’s been a long time agile person and over the last 25 years of agile, and you’ll see, like some giant company say, ‘oh, yes, we’re agile!’ And then they describe something that’s in fact the least agile thing you’ve ever heard, and they’ve just kind of sprinkled agile onto their command and control processes and say, ‘there, that’s agile.’ You’d be like, ‘no, that is not agile. Just the fact you use the word agile does not, in fact, make you agile.’

Jeffrey: And certainly I could believe he has that motivation, that dealing with the UK civil service that says because they say the word ‘systems thinking’ does not in fact mean that they are doing systems thinking that I could utterly believe. But that’s not a reason to say that agile is a bad idea, or that systems thinking is a bad idea, I think instead… But anyway, sorry, that’s the– I could imagine having that conversation with Ed, and I could justify his article a little bit if I think even that way. But look.

Squirrel: We’ll make you some chamomile tea, Jeffrey. Don’t worry. It’s all right. Hahaha.

Jeffrey: I would love to have him explain himself!

Squirrel: I’ll moderate.

Jeffrey: –To hear from him.

Squirrel: We’ve got it!

Jeffrey: Yeah, okay, fine.

Squirrel: So I think we’re now testing the theory that conflict and drama and arguments win on the internet. So if this becomes a viral podcast, we will know that what we need to do is become hot under the collar about some more topics. But let’s find out whether that’s true. And we really do welcome people who disagree with us, including Ed, to get in touch with us. And there’s a really good way to do that, which I talk about at the end of every episode, when Jeffrey calms down enough to let me do it. And that is at Agile conversations.com. Okay, Jeffrey! It’s okay! Don’t worry! Agile conversations.com is the place to find free material–

Jeffrey: No! Wait, wait, wait! Friday! Friday we’re! We’re going to be in Woking on Friday and Saturday.

Squirrel: I know! I’m gonna get there! Okay.

Jeffrey: Come to Woking! All right. I’m sorry. Go on.

Squirrel: All right. Jeffrey’s in full throated ranting mode. So, agileconversations.com is the best way, if you’re listening to this anytime after the end of September 2025, if it’s before the end of September 2025, the best place to talk to us is at CITCON, which is a conference that Jeffrey’s been running for 20 years now, I think. And, it’s in Woking, which is just outside London. So if you’re anywhere near there, both Jeffrey and I will be there. We’re intending to record drum roll, please… The 400th episode of Troubleshooting Agile with the audience participation at CITCON. And it’s a chance to come and argue with us.

Squirrel: Ed, you’ve got a free invitation. The conference is free. You know, we’d love to see you there. I’ll stand between you and Jeffrey. Don’t worry, it’ll be okay. So, please come to CITCON. You’ll find that in the show notes and also at CITCONF.com. And of course, the other way to keep in touch with us is to come back and listen next week when we’ll have that episode recorded at CITCON. And that will be when we run our next episode of Troubleshooting Agile. Thanks, Jeffrey.

Jeffrey: Thanks, Squirrel