Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Modern cities modelled as “super-cells” rather than multicellular organisms

onlinelibrary.wiley.com

Post reply on HN
Print view View original

21–30 of 54 posts

Re: Modern cities modelled as “super-cells” rather than multicellular organisms

#21

Post by rosetremiere »

siraben wrote: post #13
samatman wrote: post #11

Modeling cities as 'cellular' is a common mistake, well explored in A City is not a Tree by Christopher Alexander. http://en.bp.ntu.edu.tw/wp-content/uploads/2011/12/06-Alexan... tl;dr we tend to talk about cities in heirarchical, enclosed terms, as having "limits", "neighborhoods", "blocks", "plots", and so on. This can be modeled as a tree, in the maths and computer science sense. But it's more accurate to model th…

> But it's more accurate to model these systems as semilattices. That's the first time I've seen order theory used to describe hierarchical structures like a city. Continuing the analogy, what would the join of a nonempty finite subset of parts of a city be? Say the set was an apartment block, road and grocery store, would the join of all these concepts be the "local neighborhood" that makes up part of a town?

It feels to me this analogy is rather weak. Presumably, the point of semilattice (compared to tree) is that the hierarchy is not top-down but can have branches up and down. I'd say the algebraic structure of joins and meets is not really natural here… it's just some form of graph.

Open to counter-arguments though!

Re: Modern cities modelled as “super-cells” rather than multicellular organisms

#22

Post by dr_dshiv »

monsieurbanana wrote: post #16
telesilla wrote: post #14

Earlier quoted context omitted.

Next best thing, animated DNA : https://youtu.be/7Hk9jct2ozY

I have no words. This is fascinating. This conveys how mind blowing complex our body is better than anything I've seen before.

It's a computer based on vibrations. Every molecule has a different intrinsic oscillation!

Re: Modern cities modelled as “super-cells” rather than multicellular organisms

#23

Post by nemetroid »

rosetremiere wrote: post #21
siraben wrote: post #13

Earlier quoted context omitted.

> But it's more accurate to model these systems as semilattices. That's the first time I've seen order theory used to describe hierarchical structures like a city. Continuing the analogy, what would the join of a nonempty finite subset of parts of a city be? Say the set was an apartment block, road and grocery store, would the join of all these concepts be the "local neighborhood" that makes up part of a town?

It feels to me this analogy is rather weak. Presumably, the point of semilattice (compared to tree) is that the hierarchy is not top-down but can have branches up and down. I'd say the algebraic structure of joins and meets is not really natural here… it's just some form of graph. Open to counter-arguments though!

Are you arguing that the model proposed by the essay does not describe a semilattice? Or that the model proposed does not describe cities well?

Re: Modern cities modelled as “super-cells” rather than multicellular organisms

#24

Post by fridif »

bobthepanda wrote: post #15
samatman wrote: post #12

Earlier quoted context omitted.

New York's boroughs are another fine example. Arguably the Bay Area is something similar, and I would further argue would be much better off if it were closer in governance and self-concept to Tokyo or New York than it is in fact.

New York’s boroughs aren’t that decentralized, Manhattan is clearly the center with almost a quarter of metro area jobs and certainly not anywhere near a quarter of the metro’s population. Fun fact: Manhattan contains both the largest and fourth largest business districts in the US, Midtown and Downtown respectively.

Soon to be the 10th and 17th id imagine

Re: Modern cities modelled as “super-cells” rather than multicellular organisms

#25

Post by rosetremiere »

nemetroid wrote: post #23
rosetremiere wrote: post #21

Earlier quoted context omitted.

It feels to me this analogy is rather weak. Presumably, the point of semilattice (compared to tree) is that the hierarchy is not top-down but can have branches up and down. I'd say the algebraic structure of joins and meets is not really natural here… it's just some form of graph. Open to counter-arguments though!

Are you arguing that the model proposed by the essay does not describe a semilattice? Or that the model proposed does not describe cities well?

I was (without any better argument than my intuition) arguing that the sentence “But it's more accurate to model these systems as semilattices.” sounds weak to me. I was not making a point on the article, but only on this specific idea.

My feeling is that it's not really fruitful to describe "real life" phenomena in terms of abstract algebraic mathematical structures. It's more of a gut feeling than anything else, but I think there are two points that one can make here:

* Generally, it's better to first observe a phenomenon, understand it, and try to find a good vocabulary for it, mathematical or otherwise, rather than take an out-of-the-box fancy sounding math concept and try to make it fit your phenomenon.

* It seems as soon as you try to describe a phenomenon with an algebraic object, you'll soon get stuck trying to make sense of your functions/axioms: what's the meet? what's the join, is there a bottom element, etc. If you rather work with a more “geometric” notion (graphs, metric spaces,…) you don't have as many things to match, and you can focus on studying your objects.

To be clear, I don't have any strong point here, it's mostly gut feeling/rumblings.

Re: Modern cities modelled as “super-cells” rather than multicellular organisms

#26

Post by GuB-42 »

samatman wrote: post #12
andrewl-hn wrote: post #6

Tokyo is an example of a "multicellural" city, with its Special Wards each functioning as a separate city with only some services shared among them. The wards each have a population of a few hundred thousand people, and some take top spots in the list of biggest cities in Japan. https://en.m.wikipedia.org/wiki/List_of_cities_in_Japan The city has been run like this for decades, and it seems to work out for Tokyo pret…

New York's boroughs are another fine example. Arguably the Bay Area is something similar, and I would further argue would be much better off if it were closer in governance and self-concept to Tokyo or New York than it is in fact.

The thing is that Tokyo has no city center. The actual center is the Emperor's palace, which is mostly empty and not accessible to the public.

Other than that, each ward has its own downtown, its own suburbs. It really feels like Tokyo is not a city at all but a place where different cities decided to stick together, and administratively, it is. Shinjuku would be closest to a capital but not everything centers around it, in fact, it is not even where the "Tokyo" station is.

Re: Modern cities modelled as “super-cells” rather than multicellular organisms

#27

Post by bregma »

Not sure how this differs from Christaller's central place theory of the 1930s except the nomenclature relies on a trendy modern borrow from bioengineering.

Perhaps the authors spent more time tuning the AI they used to write the article than studying the fundamentals of urban geography?

Re: Modern cities modelled as “super-cells” rather than multicellular organisms

#28

Post by anthk »

andrewl-hn wrote: post #6

Tokyo is an example of a "multicellural" city, with its Special Wards each functioning as a separate city with only some services shared among them. The wards each have a population of a few hundred thousand people, and some take top spots in the list of biggest cities in Japan. https://en.m.wikipedia.org/wiki/List_of_cities_in_Japan The city has been run like this for decades, and it seems to work out for Tokyo pret…

Barcelona is a good and obvious example.

Re: Modern cities modelled as “super-cells” rather than multicellular organisms

#29

Post by bsedlm »

andrewl-hn wrote: post #6

Tokyo is an example of a "multicellural" city, with its Special Wards each functioning as a separate city with only some services shared among them. The wards each have a population of a few hundred thousand people, and some take top spots in the list of biggest cities in Japan. https://en.m.wikipedia.org/wiki/List_of_cities_in_Japan The city has been run like this for decades, and it seems to work out for Tokyo pret…

I suppose most big-enough metropolitan areas are like this. Mexico city is now (technically) a "state" with a few different cities (municipalities) which are nontheless seamlessly blended together.

Re: Modern cities modelled as “super-cells” rather than multicellular organisms

#30

Post by bsedlm »

rosetremiere wrote: post #25
nemetroid wrote: post #23

Earlier quoted context omitted.

Are you arguing that the model proposed by the essay does not describe a semilattice? Or that the model proposed does not describe cities well?

I was (without any better argument than my intuition) arguing that the sentence “But it's more accurate to model these systems as semilattices.” sounds weak to me. I was not making a point on the article, but only on this specific idea. My feeling is that it's not really fruitful to describe "real life" phenomena in terms of abstract algebraic mathematical structures. It's more of a gut feeling than anything else, bu…

> My feeling is that it's not really fruitful to describe "real life" phenomena in terms of abstract algebraic mathematical structures

I disagree, however I don't think you're enterily wrong. We just need a physical theory to bridge the abstract mathematics with the real world. as a starting point it occurs to me to wonder what does throwing a ball have to do with F = ma

also, the whole point of math is to do everything with such a level of abstraction that you can apply the abstractions to many things.

Post reply on HN