Dancoland | Hacker News Reader

Dancoland

alexdanco.com

68 pointsjger1527 comments

My 2 cents: 1) I really like the content - systems thinking is definitely a good tool to have, as I find it to more accurately help me understand the world around me. The post goes on to tie it into various emergent layers and such.

2) Not a huge fan of the format. Honestly it's mostly just a personal preference; I don't think there's anything inherently wrong with the whole student-teacher discussion, but to me it just comes off as self-aggrandizing. I prefer straightforward explanations, not a meandering story, but that might just be my technical side

It's essentially a Socratic dialogue. Personally I find it really refreshing to see something like this outside of Philosophy.

Socratic dialogue isn't actual dialogue. It can easily come off as condescending to explain something in a dialogue with yourself. It can easily BE condescending (tvtropes, no link: "The Watson").

This site starts off on a long winded "Socratic" lecture (and an allegory to boot) without even telling us who the author is or what he aims to convince us, in short why we should care.

And he seems to have been submitting his own writings to hacker news a lot. I'm pretty sure I saw this exact link a few days ago, in fact.

Causal statistics lecturer and old time blogger Cosma Shalizi used to start off all his blog posts with an "attention conservation notice" summary.

This author is an example of the opposite, someone who does not respect very much that the reader's attention is precious to them too.

In the hands of a master like Plato, this style is heuristic; that is, it allows the reader, indeed leads him, to discover things for himself. [...] "A master like Plato," we said - but there is no one "like" Plato. Other philosophers have attempted dialogues - for example, Cicero and Berkeley - but with little success. Their dialogues are flat, dull, almost unreadable. It is a measure of the greatness of Plato that he was able to write philosophical dialogues that, for wit, charm, and profundity are the equal of any books ever produced by anyone, on any subject. Yet it may be a sign of the inappropriateness of this style of philosophizing that no one except Plato has ever been able to handle it effectively.

Mortimer Adler

When I was doing my Open University computing degree one of the modules was on systems thinking. During the course I got told (by tutors) that I have a tendency to think more systematically than most people. So I should've been able to pass the module easily? Nope. That was the only module I abandoned; I got so angry with the way it was being taught, the assumptions and arrogance displayed by the tutors (very much "us against the reductionist world"), that I lost all interest in completing it.

At that time, the module was a required component of the computing degree. By abandoning it, I had to give up the single-subject degree. Instead I opted to take some creative writing modules and ended up with an OU open degree. Turns out I was much happier writing and creating stuff than drawing clouds and pipes and stuff.

The article seemed to me to go back and forth between profound and tendentious. To pin down the ideas precisely, or perhaps to trivialize them, here's a simple technical + mathematical example: TCP, the congestion control protocol for transmitting data packets on the Internet. Here are some ways to think about TCP:

(1) TCP's congestion control is an algorithm. It maintains a variable "congestion window", and it has rules for increasing or decreasing this variable based on the acknowledgement packets that come back.

(2) TCP's congestion control is a mechanism that results in the outcome "throughput = sqrt(2) / (RTT * sqrt(p))", where p is the packet drop probability.

(3) TCP's congestion control is a system that results in the available capacity of a network being shared between all the active flows. Specifically, it acts so as to maximize a "total utility" function, subject to network capacity constraints [1].

These three levels of description are all true, all at the same time. When you ask "why did TCP give me this much bandwidth?", the 'why' can be answered at any of the three levels, and different levels will be useful for different purposes. Level 3 is close to 'purpose' and level 1 is close to 'cause and effect', but I think those labels get confusing especially when we talk about the intention of the designer. (In the case of TCP, the level 3 purpose-style understanding of TCP only came about in the late 1990s, long after TCP was invented and deployed.)

IANA philosopher ... but these levels of explanation go back to Aristotle. I think level 1 is the "efficient cause" (it happened because of an action), level 2 is the "formal cause" (it happened so as to achieve a certain form), level 3 is the "teleological cause" (it happened to achieve a purpose).

When I read the article in the light of this three-level picture of TCP, it seems to me that the "purpose" view of a system, which I take to be level 3, is just one of several equally valid views, not more nor less important than the others.

[1] https://www.princeton.edu/~chiangm/stochastic_num.pdf

That's correct, and I think the author goes too far in implying that the efficient cause is the "wrong" way. I think what they are actually opposing is _linear_ causal thinking; after all, systems thinking is about graduating from linear to nonlinear causal reasoning.

Culturally in the West we tend to think of linear modelling as the highest truthful form (see reverence for subjects like physics, though ironically quantum physics moved away from this paradigm to a degree) which is not ideal because only a small number of subjects are amenable to such a model and most are far removed from the arenas in which we act day to day and thus would be well served with accurate models.

I've not heard of nonlinear causal reasoning. It sounds interesting. Could you suggest some reading? (A quick search throws up uninteresting hits about nonlinear models.)

[edit] this might be a good starting point:

- general overview of complex systems: https://en.wikipedia.org/wiki/Complex_system

- nonlinear dynamics (maths-heavy): https://en.wikipedia.org/wiki/Nonlinear_system

it's the underlying topic of study in systems theory in the same way that mathematics is about axiomatic models; you won't see it explicitly mentioned that much unless you go into the more pure, self-referential subdisciplines.

cybernetics focuses on things like feedback loops (where cause-and-effect runs in a loop - ie, it is not linear), systems theory works on higher-order patterns based on feedback; it spans out in many different directions (systems theory is extremely transdisciplinary) and there's a lot to it so I'm not sure what specifically I could direct you towards. "Thinking in Systems" by Donella Meadows is the typical recommendation, but I find it very diagram-focused and that's probably the least interesting part about it in my mind.

> When I read the article in the light of this three-level picture of TCP, it seems to me that the "purpose" view of a system, which I take to be level 3, is just one of several equally valid views, not more nor less important than the others.

Not only that, but you can make some pretty bad errors of thinking if you try to impose a teleology in situations where there actually isn't one.

I don't often come across a blog post that long that keeps me interested until the end. It's description of systems reminds me a lot of what I've read about cybernetics. I'd love to know what someone who has studied the topic thinks about the way it's explained.

That said, I feel like a few word games had to be played in order to reach it's conclusion; Especially with what I perceive as a somewhat conflated use of the word "purpose".

Perhaps I have to read it through again to make sure I properly understand it.

Cybernetics and systems theory are both about the same thing, which is to say nonlinear causality, so you're right to observe a similarity. The difference is primarily that cybernetics focuses on the ground up modelling (often mathematically) of casual loops whereas systems theory (and systems thinking) typically focuses on the behaviour of the system as a whole and its teleological (in this article, called purposive) behaviour, as well as things like stable states (called attractors in the literature).

To use a software analogy, they're like white box and black box analysis respectively.

One page summary diagram of this article, courtesy of the Principia Discordia:

https://principiadiscordia.com/book/59.php

Self-replying because I thought of something else after thinking about it more (it’s a pretty deep essay):

My thesis advisor once gave me a bad grade on a paper about the Gaia Hypothesis about 15 years ago, and I never really knew why. The answer I get from reading this article is that I wasn’t thinking about the steady state system as an actor in its own right. And that might be because I am inside the system so it just looks chaotic to me.

I think I got hung up on the animist implications of the theory, rather than ignoring the simple but powerful insight that any complex steady state is in some way analogous to a living organism. As I learned in SSC’s Meditations on Moloch, treating something as an actor can help reveal deeper truths.

Systems thinking is great and all but why should causality be discarded? In the example about population size, a causal graph is built that connects walking distance to population size that certainly allowed for reasonable counterfactual reasoning in that context. Yes, the variation in population size across societies cannot be adequately accounted for by causal reasoning in that same graph - you need to add and modify nodes and edges to restore counterfactual reasoning. This doesn't mean that causality is useless.

I would contend that purpose is multifaceted, not singular. Many parents (speculating, as I have not reached that life phase yet) would say that sometimes, they primarily do just clean up after their kid. We hope that we can hold multiple purposes in their head, to be human and not just a machine. Additionally purposes are subjective and not really knowable. To challenge the author who seems to argue that purpose should subsume causality, I would ask how far science could have progressed with only the former?

Not everything is part of a system. What is the purpose-driven answer to the quintessential kids question "Why is the sky blue?"

Finally, using opaque koan-like statements that are underspecified and semantically nonsensical or meaningless is a fairly daring stylistic choice (an understatement).

> What is the purpose-driven answer to the quintessential kids question "Why is the sky blue?"

The sky is blue because its "purpose" is to scatter harmful electromagnetic radiation from the sun (Raleigh scattering), a process that creates a blue hue. If the atmosphere runs out of the right kind of particles, the sky will not be blue and this loop will be broken, allowing harmful electromagnetic radiation through to the earth's surface.

It seems like the causal answer of "the sky is blue because the atmosphere scatters electromagnetic radiation from the sun and tends to scatter shorter-wavelength blue light more" is both more clear and less imbued with the moral context brought about by the use of "harmful". When you say harmful, it calls to question "harmful to whom and why?", which is important in its own right but wasn't part of the original question "why is the sky blue"? I agree that it's important to look at the context, the purpose-driven layers surrounding the "purpose of the sky being blue", but I think that in this case the causal answer is the more concise and confined.

Yeah I'm merely showing there is a way of answering this question in terms of "purpose" as defined in the article. The blue sky example clearly isn't the zinger OP intended it to be.

The point is, defining something through its purpose, is sometimes nonsensical, as you have shown (on purpose?). Sometimes - I would say most of the time - a causal definition makes a lot more sense.

Systems thinking (as described in this post) does involve quite a bit of causal thinking as well. Each of the pipes and dotted lines seem to signify a causal relationship. An increase in bugs cause an increase in food consumption.

Setting them up as opposites seems unhelpful to me.

I think the point was more that causal thinking can involve an incomplete picture, and is often merely a subset of systems thinking. At least that's what I took away. Causal thinking involves analyzing maybe part of a loop, but if you think that's the complete picture, you're missing the rest of the loop.

I totally agree with this point, but the author confusingly doesn't seem to agree:

> Cause and effect will fool you: in a world where everything is looping, there is no cause or effect. But there is purpose, and purpose moves in one direction: it is created, and it is spent.

I interpreted this to mean "cause and effect analysis helps when you look at how one step connects to the previous step, but when looking at a system you will eventually end up back at your starting point (you go around in a loop) with cause and effect analysis, whereas systems analysis sees the loop as a loop and lets you talk about the whole picture in a useful way. This is why cause and effect analysis is misleading -- because you can go through an entire loop explaining each connection and still miss the larger picture, or rather, the fact that there is a loop/purpose."

This makes me feel like a person who thought too much is re-inventing the Hagalean Dialectic, replacing the "Thesis -> Anti-thesis -> Synthesis" with "order -> disorder -> purpose". Or perhaps replacing the long arc of history

Their loop about the parent putting away the toys was not right. When a child is young a parent cleans up their child's toys in a similar way they clean their diaper: a thing of order and of safety. As a child grows, the parent teaches the child how to create order in their own life.

Chaos is all around us. (Healthy) Civilization and a Stable, Structured Family hold off this Chaos. Order is not Chaos from a different perspective. That is madness.

Just because Cause and Effect can be complicated does not mean it is not a real thing.

Thinking in terms of simple purposes (a flower's purpose is to bloom) is certainly stoic and can be useful; but that's like saying a beautiful painting is just acrylics on a canvas.

I suspect the nature of this thinking does not reflect reality and if you try to base real actions on top of it, it will end in badness.

I kept waiting for the part where he asks you to come to Our Lord and Savior Jesus Christ but like, I've got stuff to do today and only have so many hours for tedious articles evangelizing a personal point through a series of convoluted thought experiments in dialog form. It's got a few tidbits that might be worth thinking about but the HN comments have pulled out the best points.

Quick thought: reminiscent of Stewart Brand’s concept of pace layering.

——

Can’t shake the feeling that I’ve just read something profound I need to ruminate on a lot :-)

I am not a fan of the Socratic dialogue style of writing. The signal to noise ratio is low.

Reply on news.ycombinator.com