8 ms·
Systemantics by John Gall has some insightful and surprising gems that feel related to The Shirky Principle, I guess because they're both related to complex wet
by kaycebasques 3y ago
Systemantics by John Gall has some insightful and surprising gems that feel related to The Shirky Principle, I guess because they're both related to complex wetware systems.
> Complex Systems Tend To Oppose Their Own Proper Function. My favorite axiom. Your city has a problem with trash building up on the streets so it sets up a waste management company. The company starts out by collecting trash daily, but then they shift to a Tuesday-only schedule. Next, you get a notice saying that the company will no longer service your building unless you buy their standardized garbage cans (to ensure that the robotic arms on their trucks can pick them up). Eventually, the waste management union goes on strike and the trash starts piling up on the street again anyways.
> People In Systems Do Not Do What The System Says They Are Doing. The stated purpose of a king is to rule a country, but in reality they spend a lot of time fighting off usurpers.
This one feels related to the Shirky Principle:
> A System Continues To Do Its Thing, Regardless Of Need. The Selective Service System continues to require all 18-year-old male US citizens to register for the draft, even though the US hasn’t had a draft in 51 years.
https://www.biodigitaljazz.net/blog/systemantics.html https://www.biodigitaljazz.net/blog/systemantics.html (one of my many blogs)
- electrondood 3y agoThis was exactly what came to mind. Thanks for posting the reference
- kaycebasques 3y agoGreat book. I learned from its Wikipedia page that the author (John Gall) pitched 30 different publishers, got rejected by all of them, and then published it himself. It got discussed in academic papers and then the New York Times picked it up. And we're still talking about it 50 years later. Given that the central theme of the book is "systems barely work" I very much like to imagine Gall taking all of the rejections in stride because he was able to view the publishers as just another system that has its own goals and perhaps isn't really working as it should.
- MichaelZuo 3y ago> Eventually, the waste management union goes on strike and the trash starts piling up on the street again anyways. Isn't that just free association? Unless they signed a binding agreement that prohibited striking whatsoever, I can't see why they shouldn't be able to.
- brookst 3y agoI didn’t see a moral judgment against striking, just an observation that the elaborate systems set up to collect trash sometimes end up intentionally not collecting trash.
- MichaelZuo 3y agoThe union and its member's intentions sure, not the intention of the overall 'elaborate system'. Because the union and its member are only components of the larger system. i.e. A system that doesn't take into account the free association of its components is incoherent.
- brookst 3y agoI think you’re missing the original point. It was nothing about unions or free association. In fact, the point was that ALL systems become incoherent as they scale. The existence of unions and free association is built-in to the argument that systems designed to pick up trash sometimes intentionally do not pick up trash. It’s not a moral judgment that these systems are broken and therefore workers should be slaves. It is an acknowledgment that complex systems end up having to meet conflicting priorities and therefore become, as you say, incoherent.
- PaulDavisThe1st 3y agoFailure to collect trash during a strike is not "becoming incoherent as [it] scale[s]". Failure to collect trash under certain conditions is not a failure of a trash collecting system. Inability to regularly collect trash is the failure mode we're concerned about.
- matkoniecz 3y ago> even though the US hasn’t had a draft in 51 years That smuggles an assumption that US will never have draft again
- ysofunny 3y agomaybe the American way to have a draft is to fight a civil war?
- babyshake 3y agoIt seems as though there is essentially no scenario where the American public would accept a mandatory draft. Times have changed.
- fallingknife 3y agoSeemed that way in 1940 too
- hannasanarion 3y agoDid it? Every single person who was draft-eligible in 1940 was born before WW1 ended.
- lazyasciiart 3y agoI’ve done this arithmetic a few times now and I think you’re wrong: someone born the week after ww1 ended (November 1918) would have been 22 in November 1940, when the first draft registration was held for men who had reached 21. So there should be about 12 months of births in there.
- ysofunny 3y agoall I'm saying is how mandatorily drafting people to go fight a war in another continent is extremely hard whereas getting people to fight in a way that's happening near YOUR (meaning their particular case) state border is super easy, barely an inconvenience
- brookst 3y agoIt feels similar to the way agile development sometimes ossifies into rigid processes that defeat the purpose. Humans just seem wired to want to find “the solution” and the call things done.
- Aurornis 3y ago> The stated purpose of a king is to rule a country, but in reality they spend a lot of time fighting off usurpers. This quote applies to more than the people at the top. I thought office backstabbing and power plays happened at the senior management level when I was young, which left me unprepared for how much subterfuge and infighting came from ICs trying to be king of their little circle within a company. Recognizing and getting away from the people who compete by putting others down is a valuable skill in the workplace for anyone, not just the king.
- kaycebasques 3y agoYes, totally applies to everyone within the system. When I first read that I immediately started pondering how my actual work is different than my stated role. The example of the king is just the most quickly grokkable example IMO
- dreamcompiler 3y agoIMO the most valuable function a manager can perform is to quickly identify and fire toxic people. Unfortunately few managers seem able to do it.
- 121789 3y agoThese all boil down to “any collection of organisms will act the same as one organism”. It will focus on survival and self interest first. It’s an emergent property from the incentives of the (potentially well-meaning) individuals Useful for thinking through questions like “does my organization/city/etc need this new team/committee/department to exist?”
- johnchristopher 3y ago> Complex Systems Tend To Oppose Their Own Proper Function. My favorite axiom. Your city has a problem with trash building up on the streets so it sets up a waste management company. The company starts out by collecting trash daily, but then they shift to a Tuesday-only schedule. Next, you get a notice saying that the company will no longer service your building unless you buy their standardized garbage cans (to ensure that the robotic arms on their trucks can pick them up). Eventually, the waste management union goes on strike and the trash starts piling up on the street again anyways. That analogy is bit weak because strikes don't go on forever and trash pickings always resume. The city is not in the same state of perpetual trash everywhere it was in before setting up/contracting the waste management company.
- kaycebasques 3y agoI don't think Gall would disagree with you. He says "systems tend to oppose their own function". He doesn't flat-out say "systems don't work." He would probably explain its functioning in terms of these axioms: > A Simple System, Designed From Scratch, Sometimes Works. > A Complex System That Works Is Invariably Found To Have Evolved From A Simple System That Works. He might also direct your attention to this one: > The Total Amount Of Anergy In The Universe Is Constant. Gall defines anergy as the negative of energy. See also clonal anergy. “The sum total of problems facing the community has not changed. They have merely changed their form and relative importance.” You have reduced the trash on the streets, but where did you shift the anergy by reducing the "trash-on-street" issue? The real fun IMO is contemplating all the other axioms in combination with the trash collection system: > New Systems Mean New Problems. You started out with a trash problem, and now you've got a union problem. Maybe also a powerful mafia-connected monopoly problem. > The Bigger The System, The Narrower And More Specialized The Interface With Individuals. To the waste collection company I am surely just an address, 1 trash bin, 1 recycling bin, and 1 compost bin.
- renewiltord 3y agoThe purpose of the waste management company is not to manage waste, it is to gain control of waste management. That's the difference. If you have a problem, often your solution just means you have the same problem but now lack the agency to control it. Once the waste management company has control, they can then extract the majority of surplus from the problem being solved so that if your cost if the problem is unsolved is x, your cost if the problem is solved becomes x-ε. The surplus ε shrinks as the waste management company gains more control of the process. With sufficient control, ε can even go negative. The net result is that you have not created surplus for yourself. You have just found someone to transfer the surplus to.
- westurner 3y agoSystemantics (1977, 1986, 2002) > Contents: https://en.wikipedia.org/wiki/Systemantics#Contents https://en.wikipedia.org/wiki/Systemantics#Contents
- patcon 3y agoNew printing might be coming up! His wife (and heir) suggested to a friend that he might be invited to write the forward to the next one. He has cases and cases of books in his basement, and has given them to all his students for years :)
- BuyMyBitcoins 3y agoThat would be fantastic. I was hoping for a new edition.
- westurner 3y agoLinked Data! IDK YAML-LD for TinaCMS in Git, but then also Jupyter-Book because it supports notebooks and Index generation with Sphinx. Aarne-Thompson-Uther ATC character/plot story codes as Linked Data would be neat, too. https://en.wikipedia.org/wiki/Aarne%E2%80%93Thompson%E2%80%93Uther_Index https://en.wikipedia.org/wiki/Aarne%E2%80%93Thompson%E2%80%9... IIRC I sent an email to a robotics team about cataloguing metadata for supported procedures as (JSON-LD) Linked Data; there also so that it's easy to add attributes and also to revise the schema of Classes and Properties. Compared to Ctrl-F'ing a PDF copy of an ebook, Client-side JS to fuzzy search (and auto complete) over just the names of the patterns/headings in the book would be cool; and then also search metadata attributes of each. The facts in Mediawiki (Wikipedia,) infoboxes are regularly scraped by dbpedia. Wikidata is also a Wikipedia project, but with schema. Dbpedia: dbpedia.org/page/Distributed_algorithm: https://dbpedia.org/page/Distributed_algorithm https://dbpedia.org/page/Distributed_algorithm dbpedia.org/resource/Category:Distributed_algorithms: http://dbpedia.org/resource/Category:Distributed_algorithms http://dbpedia.org/resource/Category:Distributed_algorithms dbpedia.org/page/Category:Anti-patterns: https://dbpedia.org/page/Category:Anti-patterns https://dbpedia.org/page/Category:Anti-patterns IIRC there used to be a longer list of {software, and project management} antipatterns on wikipedia? It may have been unfortunately and sort of tragically removed due to being original research without citations. Anti-pattern: https://en.wikipedia.org/wiki/Anti-pattern https://en.wikipedia.org/wiki/Anti-pattern Also there's an Antipatterns catalog wiki: https://wiki.c2.com/?AntiPatternsCatalog https://wiki.c2.com/?AntiPatternsCatalog There are probably more useful systems patterns to be mined from: the Fowler patterns books like "Patterns of Distributed Systems (2022)" [1] and "Patterns of Enterprise Architecture", Lamport's "Concurrency: The Works of Leslie Lamport", Leslie Valient's Distributed Systems work, [1] https://news.ycombinator.com/item?id=38234304 https://news.ycombinator.com/item?id=38234304 Perhaps there's also general systems theory insight to be gained from limits and failures in [classical and quantum] Universal Function Approximation; general AL/ML limits and Systemantics. Universal approximation theorem; simulacra: https://en.wikipedia.org/wiki/Universal_approximation_theorem https://en.wikipedia.org/wiki/Universal_approximation_theore...
- cratermoon 3y agoGall mentions the example of an organization founded to conquer polio. With that disease all but eradicated, the foundation almost collapsed, but instead changed the goal to conquer genetic defects. Very little changed about what the organization did, and the new goal is one that isn't likely to ever be completely achieved.
- matthewdgreen 3y agoThat doesn’t seem like an entirely bad thing, though. Once you have an organization built up to solve a problem, shifting those resources into solving another problem is a reasonable next step. This is not the same as TFA, which talks about organizations preventing the problem from being solved.
- cratermoon 3y ago> That doesn’t seem like an entirely bad thing, though. Not at all. It's an example of a good pivot. Probably Gall thought it rare enough to mention it in a book otherwise full of wisdom about how systems go wrong.
- bigstrat2003 3y agoI disagree. When you've solved the problem the best thing to do is to celebrate, give resources back, and then organize to solve another problem if you want. Switching missions is a bait and switch for your supporters and should be avoided.
- matthewdgreen 3y agoOrganizations are expensive and hard to build. If you liquidate an organization, you might free up some modest amount of cash in a bank account (and maybe some furniture, office supplies, etc.) but the cost of rebuilding that same organization will be vastly higher than the money freed up. It’s like tearing the copper out of a working industrial air conditioner and selling it for scrap: a huge waste and a loss of value in absolute terms. Perfectly reasonable to give back money to recent donors who opt out and raise it over again, though.
- fallingknife 3y agoSelective service is a terrible example because it is designed to draft for an unknown future war, not for a current one.
- dweinus 3y ago> Your city has a problem with trash ...doesn't that entire problem rest on the fact that the unstated goal is to pick up the trash at the lowest cost labor will bear? That yields service reduction, automation, and labor disputes. When you look at it that way, the organization is seeking the goal, not fighting it. I think there is truth in the general idea, but a correlary that no one sees value in continuing to solve a problem that has been solved once. They always imagine it will get cheaper, meanwhile those who create expertise in it see themselves as more valuable with time, and that creates tension.