Hacker News new | comments | ask | show | jobs | submit login
How to Be a Systems Thinker (edge.org)
283 points by raleighm 10 months ago | hide | past | web | favorite | 82 comments

A great book for this is Thinking in Systems by Donella Meadows. She has many illustrative examples that really help get the points across. Coming from an electrical engineering background, I found her discussion around the interactions between positive and negative loops to be very interesting.


I tried reading this book several years ago and bounced right off. Given the multiple recommendations, I'll give it a second try.

You can always start with the summary given at the end of the book (IIRC), and maybe that'll be enough to trigger your interest and motivates your going through at least some of the chapters.

A possible alternative might be The Fifth Discipline.

I found it a bit easier to process.

Agreed. This is a nice primer.

Hmm, someone claims that she invented Malthusianism. She must have a time travel system....

"Limits to Growth―the first book to show the consequences of unchecked growth on a finite planet"

"How do you deal with ignorance? I don’t mean how do you shut ignorance out. Rather, how do you deal with an awareness of what you don’t know, and you don’t know how to know, in dealing with a particular problem?"

I don't really like the rambling tone of the article which never really addresses it's central premise-questions.

Look, systems thinking is about thinking. Thinking is about how you were trained to think. To become a systems thinker, learn how to think better, how to think differently, how to think critically, and how to think logically. Now self-examine, correct problems, and repeat loop.

For example, the thing many elite schools teach that most others don't is the trivium and quadrivium approach to thinking.

Another trick is less tangible. Find brilliant people and learn from them. Learn what they do wrong, not just what they do right.

Finally, as a sysadmin, if you learn to think in a flowchart about just about any problem you will be fine... but the trick to that is to be specific about what the problem is.

agree there are better authors to learn how to think: D. Kahneman, A. Tversky, D. Hofstaedter, D. Dennet, K Popper, B. Russell, K. Goedel, N. Taelb, ... if you are in this same echo-chamber as me I'd love to hear who I forgot

My experience here is if you start with one of the above names (and are a fan of logic), you most likely end up reading all of them. (But maybe it's just me idk?)

Wittgenstein, Gerry Weinberg, Korzybsky, and The Systems Bible by John Gant. Gregory Bateson.

Meadows and Forester.

That should be "Forrester".

Systems thinking is not just about thinking. It’s about understanding causality and feedback loops. Sometimes these things are hard to understand intuitively but can be better grasped with mathematics and network theory.

The academic approach seems to portray systems as static, inflexible, snapshots in time.

However, the formation of systems occurs over time; so the practical art to Systems Management involves understanding the rate of change, flexibility of inter-dependencies, and sustaining momentum.

What a great last paragraph:

"The tragedy of the cybernetic revolution, which had two phases, the computer science side and the systems theory side, has been the neglect of the systems theory side of it. We chose marketable gadgets in preference to a deeper understanding of the world we live in."

Yes! I have been bewildered as to why Cybernetics has been so woefully neglected.

"Introduction to Cybernetics" is available as a free PDF. (Linked from this page: http://pespmc1.vub.ac.be/ASHBBOOK.html )

Cybernetics is formally beyond logic. It deals with circular loops in causality in the most general yet concrete sense.

One way I try to describe it (to an audience like HN readers) is that Information Theory is about making the real world behave like symbols, while Cybernetics is about making symbols behave like the real world.

This sounds very interesting. I don't think I ever encountered Cybernetics in college nor in all these years as a professional software engineer. Any more resources apart from the linked pdf that you would recommend, perhaps tailored to someone with a CS background? (I'm reading through the pdf, just wanted to see if there were other treatments on this topic)

"Introduction to Cybernetics" is technical and very much in the spirit of CS. A lot of what gets called "cybernetics" tends to be kind of "fluffy" and I personally avoid that stuff. I'd recommend Norbert Wiener's "Cybernetics: Or Control and Communication in the Animal and the Machine".



And "Behavior, Purpose and Teleology" by Rosenblueth, Wiener and Bigelow:


Interesting. The people who know how to make marketable gadgets must understand systems theory in order to sell it.

It's interesting because it's about culture and culture is a system of ideas, rituals and behaviours.

We are living in a cybernetic culture right now, so what's next?

That's what we need to understand. What do we want?

>... we’re all seeing is that a lot of work that has been done to enable international cooperation ... pulled apart. We’re seeing the progress we thought had been made in this country in race relations being reversed. We’re seeing the partial breakup ... of a united Europe...

> One of the problems when you bring technology into a new area is that it forces you to oversimplify.

It is fortunate that we have thinkers such as this to simplify the world for us.

Should we pine for the united Europe of Rome? Of the Reich?

Was it evil to tear down those cooperative societies?

REG: All right, but apart from the sanitation, the medicine, education, wine, public order, irrigation, roads, a fresh water system, and public health, what have the Romans ever done for us?

XERXES: Brought peace.

REG: Oh. Peace? Shut up!

-- Life of Brian

Yeah, I noticed that. Then she went on to talk about how she was asked to write about systems thinking even though she knows nothing about it, and wrote a preface based on a few minutes googling.

I admit I stopped reading at that point. Bog standard Marxist thought in academic circles? Starting an article supposed to be about systems thinking with a noisy announcement of tribal loyalty? Yawn.

Well, we give the Charlemagne Prize, despite his unification efforts not being exactly democratic.

That said, I think the author could argue that by mentioning "progress in race relations" they are disqualifying the Third Reich.

I jarred a bit at her use of the word "metaphor" to describe understanding, and upon reflect I think I understand why I reacted that way. Metaphors lack the requisite understanding of an analogy in that they apply at best partially to artifacts of the intended reference. An analogy has to match, or at least rhyme, with the logic of a system in order to apply.

There is a reason that analog computing mattered, and why no one seriously focuses on metaphoric computing (except, you know, as a metaphor for something that isn't computing). I suspect that metaphors are the tools of cargo cults while analogies are the tools used by those that actually understand a system in order to convey it to those that do not.

> I suspect that metaphors are the tools of cargo cults

Perhaps I'm misunderstanding, but are you using some much more restrictive definition of 'metaphor'?

For example, as far as I can see, one can't escape metaphor because they're deeply embedded in the very language you're using right now.

The most basic metaphors are ones that refer to the body or the physical world. Jar [1] originally refers to a 'screeching sound'. Understand probably from the literal sense to 'stand between' [2]. Compute derives from 'com' (together) + 'putare' (reckon, settle an account) [3]; the latter root originally comes from Indo-European 'to prune'. Almost all vocabulary is in fact a form of extended metaphor; in fact, it cannot be any other way, because to describe a new thing requires one to compare and contrast it to existing things.

You yourself are making a metaphor when you say that an analogy has to 'at least rhyme' — a metaphor for poetic sound similarity to stand for a sort of logical similarity.

The distinction between an analogy, where one explicitly calls out the comparison ('like', 'as'), and a metaphor, where the relationship is taken implicitly, is much blurrier than it seems. All our language and thinking is saturated with metaphors at every single level, to the point where they seem to be taken for granted and disappear.

IMO, the problem is not metaphors themselves, which are unavoidable; it's picking wrong metaphors (which is really using the wrong language) that lead us astray.

[1] https://www.etymonline.com/word/jar

[2] https://www.etymonline.com/word/understand

[3] https://www.etymonline.com/word/compute

"at least to rhyme" is very specifically a metaphor in that it does not require systemic logical matching... I don't literally require rhyming or a logical system that is comparable. It's also utterly useless for _system thinking_, which is the author's subject matter, precisely because it is at best loosely related.

An analogy, or analog, is a much deeper match in that means to transfer the understanding and knowledge of an existing system to a new target system that closely parallels that understanding and knowledge. Analog computing was literally "creating a model of the problem" that closely matched the real problem.

Agreed that metaphors are not inherently bad, but the dependence on metaphors rather than analogies is symptomatic of a lack of depth in understanding when attempting to convey knowledge.

To get a better understanding of her use of the word "metaphor" I recommend the book Metaphors We Live By by George Lakoff. It makes the argument that the language (specifically the concepts we build using the language) we use is metaphorical. It goes on to show how wholesome the bind between metaphors and concepts is.

From reading the book, if someone says that "In a complex world we perceive, think, understand and communicate metaphorically.", I get what they mean. It occurs to me that her use of the word "metaphor" follows from such a comprehension.

>I suspect that metaphors are the tools of cargo cults while analogies are the tools used by those that actually understand a system in order to convey it to those that do not.

Are the metaphors actual tools, or more like metaphorical tools?

Crappy tools are still tools.

It's a metaphor, metaphors aren't literally physical work implements.

And tools are not limited to physical implements.

They aren't, we can easily extend the concept by using it as metaphor.

Metaphor isn't necessarily vague. You can have extended metaphors, which can have surprising and informing depth. It can even be an isomorphism.

e.g. the desktop metaphor

The "desktop metaphor" is a conceit, in the literary sense, that provides a general structure to a presentation but it isn't truly useful as a means of conveying the logic or expectations of a system in the way that an analogy provides.

An aspect that does is the position of a mouse on a desktop with the position of a cursor on the screen.

Except that the "desktop metaphor" stops being a "desktop" with the introduction of a cursor and mouse. Actual desktops are manipulated directly by users... and desktops don't allow you to move your hand all the way to the right and magically move to a new clean desk (in the case of virtual windows). There's nothing about a desktop that infers the behaviors described which is why it's crappy for conveying the actual behaviors of current UI... hence the limited nature of metaphor.

Given an appropriate physical implementation of a desk, more aspects the desktop metaphor could certainly be implemented. Typically metaphors allude to the platonic ideal of something and not any one concrete instance of a given subject.

I'm not sure that metaphors allude to Platonic ideals but rather (to stick with Plato) are closer to the shadows on the cave wall.

>. It’s important to be aware of it, to realize that there are limits to what we can do with AI. It’s great for computation and arithmetic, and it saves huge amounts of labor. It seems to me that it lacks humility, lacks imagination, and lacks humor. It doesn’t mean you can’t bring those things into your interactions with your devices, particularly, in communicating with other human beings. But it does mean that elements of intelligence and wisdom—I like the word wisdom, because it's more multi-dimensional—are going to be lacking.

This I think is a very salient point. What is missing in the field at the moment to truly advance is a theory of 'agency' or how intelligent systems can be trained to venture out into the world and develop their own models, set their own incentives and continuously interact with their environment and develop their own 'games' from unstructured data rather than optimising towards some fixed goal.

There are actually a lot of great materials available on Systems Thinking, here are a few of our favourites:


- Introduction to Systems Thinking (https://thesystemsthinker.com/introduction-to-systems-thinki...)

- Tools of a Systems Thinker (https://medium.com/disruptive-design/tools-for-systems-think...)

- The Mythical Leverage Point (https://blog.kumu.io/the-mythical-leverage-point-d582ce4b8b4...)


- Peter Senge Introduction to Systems Thinking (https://www.youtube.com/watch?v=eXdzKBWDraM)

- I Used To Be A Systems Thinker (https://www.youtube.com/watch?v=3Ymt_TbNNwE)

- The (Failed) Promised of Systems Thinking (https://www.youtube.com/watch?time_continue=614&v=aelqgvFXGi...)

- Systems Practice Mindsets (https://vimeo.com/212281432)


- +Acumen Systems Practice Course (https://www.plusacumen.org/courses/systems-practice)


- Thinking In Systems: A Primer by Donella Meadows (https://www.amazon.com/Thinking-Systems-Donella-H-Meadows/dp...)

- Fifth Discipline Fieldbook (https://www.amazon.com/Fifth-Discipline-Fieldbook-Strategies...)

- Systems Thinking for Social Change (https://www.amazon.com/Systems-Thinking-Social-Change-Conseq...)

Articles on Leading Systems Change:

- Dawn of Systems Leadership (https://ssir.org/articles/entry/the_dawn_of_system_leadershi...)

- Acting and Thinking Systemically (https://thesystemsthinker.com/acting-and-thinking-systemical...)

- Transforming the Systems Movement (https://thesystemsthinker.com/transforming-the-systems-movem...)

Relevant Tools and Websites:

- Kumu (https://kumu.io) - Web-based tool for building interactive system maps.

- The Systems Thinker (https://thesystemsthinker.com/) - Complete library of all "The Systems Thinker" publications over the past 30 years

- Loopy (http://ncase.me/loopy/) - An playground for building interactive systems maps.

(Disclosure: I'm a Developer at Kumu)

- "General Systemantics" by John Gall ( https://en.wikipedia.org/wiki/Systemantics ) It's humorous but I'm being serious.

He has such chestnuts as: "A complex system that works is invariably found to have evolved from a simple system that works."

Love John Gall's approach to systems. There's actually a section in Kumu's manifesto that quotes Gall's law:

    A complex system designed from scratch never works
    and cannot be patched up to make it work. You have
    to start over with a working simple system.

Great post, but lofty.

Here is an example of a relatable situtational system thinking in the very small:


I havent read much on the theories, but part of my write was also inspired by two books and hence why I am responding to help others. Pragmatic Thinking and Learning by Andy Hunt and The Fifth Discipline by Peter Senge helped me with this type of thinking.

A few things key concepts. Organization of related systems and their interaction patterns. But a key, that you are also apart of the system as you interact.

For instance, the causal link of bringing sweet smelling dry erase markers. You could shift the whole system of interview rating based on memory of smell on that sweet expo choco mint as you erase your complex mistakes away.

Or they could think that you're just crazy.

I want to like the idea but articles like this one are just intellectual drivel to me. People are systems thinker by nature I believe. If there are negative side effects to their actions then it's rather because they don't mind. A company, for instance, cares about environmental side effects only as much as it needs to in order to keep going. But that does not mean they don't see, say, their supply chain as a system.

I'd say thinking in systems is more of an art than a rational thing. You have to do it as often as possible to get better at it but there isn't much you really need to know. Sit down, think really hard for 2 minutes, rest a bit and repeat the process as often as necessary.

"Oh, you're a theoretical physicist? I find theoretical physics is just thinking. Think hard, and repeat!"

There's more to both the study of systems and the mindset than "think". Knowledge of a subject is knowledge of differences and distinctions- reductionism doesn't mean you know something, often quite the opposite.

It took many of us a long time to better think about complex systems, especially those systems we grew up as a part of. There are a number of useful mental patterns that have helped me in my limited journey (eg OODA, feedback mechanisms from cybernetics) and it's not clear to me I'd have recognized the concepts were valuable enough to name on my own.

Article doesn't define it, but "systems thinking" is an actual thing: https://wikipedia.org/wiki/Systems_theory

Oh thank you thank you. At first I thought was justing thinking in the systems engineering area (which it is partly).

And then I grew more mystified as the article didn't touch on that.

So I suspected that it was just waffle which that wikipedia article addresses:

> [criticised as] ... said to be nothing more than an admonishment to attend to things in a holistic way.

Which I think is actually right. Too high level to be actionable or definable.

Systems Thinking has been a huge buzzword in the Coursera course catalog in recent years (that's where I first heard the term called out as a special thing outside of just "how we learn science"): https://www.google.com/search?q=coursera+systems+thinking gets many hits.

I've been seeing a lot of threads/articles about systems thinking come up on HN lately. As a developer, why is it important to learn "systems thinking"? How will it make me a better developer?

> How will it make me a better developer?

Well, focusing a component of the system (software development) while neglecting to consider how that component interacts with the whole system (society and ecology), is exactly the sort of myopic thinking that systems thinking proclaims to remedy.

I find systems thinking an interesting subject but honestly I think it is more useful for people higher up (PMs, etc) and/or people whose primary work is more along the lines of strategizing and evaluation more than implementation

Seems to me like category theory for mathematics, but a more generalized one that could be applied to things in life. Sounds cool

What are good resources to help start thinking in systems when designing software?

It may seem glib, but the best resource to spur one to think systemically is a three year-old child. Their favorite word seems to be "why?" and they will keep asking it until you either reach an abstraction they accept, you get tired and concede you don't know by saying "because I said so," or they get tired and find something else. The third option is rare in my experience.

I found this paragraph especially thought-provoking: "Americans are inclined to talk about the "war against drugs," or the "war against poverty," or the "war against cancer," without questioning whether "war" is an appropriate metaphor. It’s a way of talking about complexity, but if it doesn’t fit, it will cause you to make errors in how you deal with your problems. The war on poverty failed partly because poverty is not something you can defeat, and that makes warfare an inappropriate metaphor. The same is true with the war on drugs, which has gotten us into some ugly situations. "

I think war is used more in the sense of a "mission" in such cases. People see the single-mindedness, cooperation and the spirit of sacrifice that happen during a war, and wonder if why we cant bring that to bear against the other challenges we face.

At work, we sometimes set up a "war-room" during production downtime - again the logic is the same : we need to work with single-minded focus on this one problem and all else is secondary to getting the system up again.

Compare "jihad" or "Kampf", terms used in Arabic and German for a mission or struggle against some thing or for a cause (though both have also acquired negative connotations).

Actually the term "war" in the contexts of drugs and poverty is appropriate since they both depend upon force as the means to achieve their goals as they depend upon government action.

It is only appropriate if you cannot conceive government actions that have nothing to do with force :)

And what actions would those be that do not, eventually, rely upon force?

All laws rely ultimately on force.

Would you call policing a 'war on crime'? The enforcement of contracts a 'war on fraud'? The collection of taxes a 'war on free-riding'?

The deficiency of 'war' as the correct metaphor here is that war implies much more than simply 'force'; war has a particularly structure, in part:

1. It is a conflict between distinct groups

2. It is the exception and can be distinguished against 'peace'

3. Victory, or defeat, is possible (in ancient times, where warfare was common and vicious, the literal elimination of entire polities was very common)

None of these things are true of the 'war on poverty' or 'war on cancer', for example.

Unfortunately we do hear the phrase "war on crime (usually meaning violence or drugs)," but policing is the enforcement of all laws and thus lacks the targeting required to be a "war" in a meaningful sense. The same with contracts and taxes, although both have suffered their own specific "war" monikers on particular behaviors in the past.

A war does indeed have a structure, but I'd disagree with your list a bit. Your point 1 is correct, but "groups" is a fungible category and we've certainly seen wars that included the attempted destruction of both groups of people and concepts (history, religion, ideology, etc). 2. Unfortunately history argues that peace is the exception rather than war. There are plenty of wars occurring right now in which we have no participation, nevermind those that we do. 3. Plenty of wars ended with no victory or defeat, and it's certainly arguable that in some cases neither was realistically possible.

I specifically left out the War on Cancer because a good portion of it was under the aegis of private parties, not that it much mattered when it came to the failure of results. They originators believed that Cancer was a single entity that one muster all efforts against... they were wrong.

That the wars waged were wrong-headed in both conception and means is pretty obvious, the means simply ensured that the consequences would be much closer to actual war.

Whether or not things eventually rely on force seems like more of a metaphysical question than a political one; the resources spent on the perpetuation of the military-industrial complex could have been (could still be) diverted towards raising the living standards of the poorest, ensuring a basic security net for all, providing decommodified healthcare and education. It is interesting to think about why this choice wasn't made, why it doesn't get made.

Never-mind the military-industrial complex. A nation-state is in itself a monopoly on sanctioned-violence expressed through your local police department or standing Army. Property rights, human rights, taxes, and everything funded by government revenue is only possible because of the guys with the guns projecting government Force.

This isn't a metaphysical question, this is literally how modern governments are able to exist and operate at scale and how your dollar bills or local equivalent have any value at all.

It isn't metaphysical, it is purely a practical observation. How do you propose to "raising living standards" and all the other offerings using government but not using force?

The relationship between government activities and government use or threat of force is highly nonlinear.

The biggest chunk of government use or threat of violence is in property rights enforcement. Anything on top of that is mostly peanuts.

For example, if a local government raises property taxes by 5 percentage points in order to get the funds for better schools, there's likely no additional use of force at all. So it's absolutely possible to raise living standards without use of force for all practical purpose.

A reasonable libertarian who worries about use of force (but I'm aware that such people are so rare they may as well not exist) should be in favour of such a policy, because it increases the ratio of "quality of life per use of force".

A reasonable person, regardless of political affiliation, understands the consequences of the policies they support. If you believe that raising property taxes has no use of force you should attend a sheriff's sale of property seized due to lack of property tax payments.

It is not the linearity of the USE of force but the threat that matters. I support any number of government actions knowing full well that those actions very well can result in the use of force. I understand and accept those consequences as part of my moral duties rather than try and hide from them.

By transferring government resources from the perpetuation of force (military industrial complex/prison archipelago) towards the social expenses mentioned above. This would involve a diminution of force, and an increase in cooperative social relations.

Let's be clear, we're not disagreeing on policies we're disagreeing on what the behavior of government is in enacting policies. Government does not have resources it did not acquire through laws... which are by necessity enforced. Say "no" to the IRS and see how things work out.

A diminution of force is not the absence of force and you simply can't have a government act without the consequence of enforcing its actions. This isn't "cooperative" since that requires voluntary consent of all parties. I may support any number of government actions but I do not hide behind euphemisms to convince myself that those actions carry with them the very real potential and actual consequence of force.

I think you're not catching what fallous is saying, because s/he's using libertarian vernacular. Libertarians view taxation as force, because if you don't pay the tax the police can forcefully take your property/valuables. IMO this view is shortsighted and doesn't properly assess the advantages to generally granting the government a monopoly on the use of force.

You're mistaking my pointing out the means as a critique of the ends. If you do not pay your taxes not only can the police take your property but they can and will imprison your person. That is the reality of the situation and I support the collection of taxes (and the enforcement required) with the full understanding of what my support entails. I don't try and drape fig leaves around it.

Do you use "force" to grow a garden?

Gardening is often proposed as an alternative metaphor to the more common metaphors of war and sports.

The defining characteristic of a lot of garden styles is meticulous curation; anything undesirable is torn from the ground and discarded.

Only in combination with setting up and maintaining an environment designed for the plants that are desirable. Lots of supporting action, not just weeding out the bad parts.

Providing subsidized college education, for example?

And that money comes from where?

Citizens, based upon the social contract.

Hobbes recognized that the social contract is enforced by the sovereign, otherwise we have anarchy. That good is derived from such a contract does not mean that it is absent force... it's necessary. It is absurdity is to try and hide from that fact just because we like the initiative or results.

The specific citizens in that case are those that run banks, and contracts provided by banks are backed by governmental laws.

This is an illustration of the point the author was making in the first place. You have to pick your metaphors carefully because they come with secondary aspects you tend to forget: in addition to governmental use of force, war introduces "enemies" and zero sum thinking into the decision making process, sometimes subconsciously.

It's also worth noting that war is necessarily an admission of failure -- failure of all other means to resolve a problem.

Indeed. By choosing to conceptualize the effort as "war" it necessarily implies means and goals.

I'm afraid I was arguing against the use of the war metaphor...

As was I.

Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact