Next up in #PapersInSystems: we’ll be discussing Crafting Conceptual Integrity: ch 2 of Learning Systems Thinking
Discussion will be led by (the author!) @diana
Monday, February 10, 1-2pm Eastern Time
Free, but enroll for access:
https://ti.to/bredemeyer/montalion_conceptualintegrity
Scanning for “appetizer” quotes —
‘Technical debt is how we describe the lack of cohesion in the system.’
— I hear @diana ’s voice in my mind, especially at parts like the last phrase in:
‘Donella Meadows defines systems thinking as how “parts together produce an effect that is different from the effect of each part on its own.”
Relationships produce effects. Software becomes a system of software when “parts together” achieve something that could not exist without the “together” part.’
“For many software professionals, the biggest mindshift when thinking in systems is also the most important one: software systems are sociotechnical. When I say our systems are sociotechnical, I mean that our thinking, behaviors, and communication patterns are inextricable from the software systems we produce.”
— @diana
“The states change over time nonlinearly. The basic model hides the complexity in most real-life systems because we rarely have one goal at a time. Also, you can’t control everything happening outside the state box while the discrepancy diminishes. While you are changing a system, the world is changing around it.”
— @diana
I keep being reminded, reading the chapter and further into the book, that Diana puts even the things we know, in just enough of a new light that suddenly we see something new, or it gains new emphasis, casting new light on something else we now understand differently. Or can express differently.
[that is, the book is titled “learning systems thinking” but is not a survey book, though it draws on Meadows and others, and lends its own insights to the field]
Divide and conquer, it turns out, is not about effective — resilient — complex systems.
@RuthMalan @diana I had no idea she was on here! I’ve got the book but have yet to read it. Looking forward to it!
@RuthMalan @diana where I feel mathematical systems or physical systems can be changed very slowly ("adiabatically", e.g. moving a nonlinear system towards a design goal, reprojecting the parameters towards optimality with each tiny step), sociotechnical systems are a bit more quantized - you cannot change a process with imperceptible steps, people will just fall back into the original one if the step is too small
@RuthMalan I learned from @tottinge that we have divide-and-conquer backwards. In warfare, you separate the enemy, then pick them off with overwhelming forces. We have made ourselves and (our systems?) the enemy.
@qcoding @RuthMalan @tottinge I don't know of another meaning for that expression. What are the other ones?
@fanf42 @RuthMalan In a comment @tottinge wrote 3 years ago:
When you "divide and conquer" it is the divided force that is conquered. The idea is to break up the larger force so that you can overwhelm and defeat the smaller pieces of the large force.
When people say Divide and Conquer in our space, they usually mean that they are going to divide their own forces to the maximum degree possible (individuals) and somehow by dividing their OWN forces, win.
"Our space ==?" TDD? Systemic?
OK, never heard of that meaning. It seems extremely far fetched from the original saying, almost a misinterpretation.
Thank you for the explanation.
@RuthMalan @qcoding oh, decomposing a complicated problem in simpler bit is something I can understand. But I just never heard "divide and conquer" for that (and in that case, it's the problem that is divided, so it seems to apply somehow in the expected way)
@RuthMalan @fanf42 @qcoding That's the proper use. That's not what we see in the wild.
@fanf42 @qcoding @RuthMalan "Our Space" == "Corporate Software Development" if Jon means what I have seen the most often. This idea of "divide ourselves" is pretty ubiquitous, because people have (nothing but) faith that solo tasks in parallel is the same as maximum efficiency.
That's a common hopelessly naive misunderstanding of the term, and also of Amdahl's law.
@fanf42 @qcoding @RuthMalan In software, it's become common to say "let's divide and conquer" meaning "let's all split up and work individually, and integrate it later when all of us are done."
You're not supposed to divide your own forces and be conquered.
The point is to have more force on your side by nibbling away parts of the enemy.
Properly, that would be "what part of that can we solve together, today?"
@RuthMalan @diana not sure why* I'm thinking to that, but I once read a blog post about a research social framework helping understand rapid shift in human social contracts with "fall to autocracies". Not saying it like that, but it presents two feedback loops for two hulan mindset: the first one wants order, perceived predictability (absence of surprise), belonging (oneness), and saneness. The other wants universality, diversity, and don't care that much for oneness (not at the community scale).
The former will accept violent and rapid changes (ie time based disruption) when looking for space based shrinking of perceived diversity.
The second loves large space based diversity. He is frightened by rapid changes / fight against them to ensure time base continuity.
W-en the second group gradually complexify the world, the first feedback loop grows more potent until breaking equilibrium.
For me, it's one of the first useful framework to explain why people vote for authoritarian regimes even if it's bad for them:
https://hopenothate.org.uk/2020/11/01/authoritarianism/