Why so few Venice-like cities?

So asks Sam Wilson on Twitter.  You can parse this question a few different ways, but I take it to be about location theory and, indirectly, topology.  That is, why are there so few cities with multiple water connections between different independent nodes?

First we might check the historical premise.  Stockholm used to be called “Venice of the North,” a variety of Dutch cities are based on canals, check out the (now largely defunct) Marsh Arabs from southern Iraq, plus a variety of arrangements in southeast Asia over the centuries, such as kelong, or the Pang uk of Hong Kong.  There is a long history of stilt houses on water, including but not restricted to Chiloe, Chile and Nipa huts in the Philippines, some of which are designed for wet areas.

But I digress.

In any case, many cities seem to have a much smaller number of nodes connecting to the water.  Check out Duluth, Minnesota.  There is a dominant port, with most local connections to that port running overland rather than by water.   And if a goods shipment is headed into Duluth, you pretty much know where it will arrive.  The same is true for Rotterdam, as no one is angling to reach those inner, smaller canals of Dordrecht with their bananas.

The greater the anonymity of exchange, and the greater the distance involved, the stronger is the role of a formal port as a centralized supplier of trust and also buyer-seller coordination.  That will imply a small number of water nodes, all the more so as globalization and specialization proceed.

But if you are a Marsh Arab wishing to trade some rice for some coriander with your neighbor, and the next day lend a sowing needle in return for some gossip, to a different neighbor, such idiosyncratic bilateral yet multiple exchange nodes and networks may work pretty well.  You can think of multiple node cities as doing a great deal to enable non-regular, non-standardized barter, but that becomes less valuable with economic growth.

A related approach is to ask when it is more efficient to settle nearly contiguous islands, as opposed to dry land next to water.  In earlier times non-monopolized access to fish, trade, and water transport were main reasons, plus Malthusian conditions elsewhere, such as in the resource-sparse, low-yield Veneto.  Venice was settled long before a mix of accident and leadership skill led to it becoming a primary conduit between East and West and thus a wealthy and powerful commercial republic.

For a variety of propositions about monetary economics, there are corresponding propositions about topology and location theory, the trick is to see them.



In general, where you have a lot of low-lying islands, such as Venice, the water channels tend to be shallow as well (unless dredged), which makes it hard for deep-draft oceangoing vessels to dock. Moreover, low-lying islands are vulnerable to flooding by storms.

In contrast, the city of San Francisco has a famously useful northeast coastline, where deep water comes up very close to solid, non-floodable city land. This helps answer the question of why foggy, clammy San Francisco is the civic center of the San Francisco Bay Area instead of, say, sunnier Oakland or San Jose: what's now downtown San Francisco was the easiest place in the Bay Area in 1849 to quickly build piers for ships coming around the Horn.

Mexico City was built on a shallow lake before 1325. By 1519 it was the largest city in the world.

The islands were artificial. You bundled logs together, set them in a loop, and dredged mud from outside into the middle until you had an island (a chinampa) and a canal around it. The main streets of the old commercial centers used to all be canals and chinampas.

After the big flood in 1460, Mexico started building a series of dikes, tunnels, pumps, diversions, and other multi-billion dollar water projects. For six centuries continuous new and upgraded water projects have mostly kept Mexico from flooding.

By the 1990s Mexico was the largest city in the world again (or second). All the canals had been filled in, but their traces and geometry were fixed in the street grid. Canals stink -- even in Venice -- and attract flies and mosquitoes and require expensive maintenance. You need bridges to cross them. Streets have been better for local transportation, faster, and cheaper for more than two centuries now.

The water engineers told us that covering the last quarter of the old canal city in streets would flood the whole city. Water percolating down of the Águila cordillera of volcanoes needs a permeable place to sink down into the water table. So the canals of Xochimilco were preserved. Five miles from dead center downtown of the twenty-two million megalopolis you can find fields of corn, cattle grazing, decorative flowers growing, garden nurseries, organic arugula, hog wallows, open water, skull racing, and rural farmers' straw huts. Hundreds of miles of canals weave an archipelago with pure rural peace and semi urban garden islands and Venetian densely packed city islands where no motorcar could fit.

Unlike Venice, it has not all been turned into a tourist trap. Middle class and working class families make Xochimilco a working landscape. But it survives by ecological necessity. If it were possible to pave it safely, the Mexicans would have done it.

Canals stink — even in Venice — and attract flies and mosquitoes

That's because the canals double as Venice's sewage system. Without tidal flow, the place would be a depopulating, disease-ridden cesspool.

Which is a big reason there are rather few cities like Venice: sewage is a huge problem. Even coastal cities with the wrong water flow can suffer: Chicago did not take off until the Sanitary Canal carried its effluvium far away rather than it backwashing into its water supply from Lake Michigan.

This is fascinating. I missed Xochimilco on my visit to Mexico City, but am making a note in case I ever go back.

Mexico City was founded for the same reason as Venice, as a place for refugees who had lost the wars for the surrounding area to go to and survive. Through smarts and luck, the refugees and their descendents than built an empire.

You can print a map of the Centro of Xochimilco and walk to many interesting islands. Look on Google Earth or Google Maps Satellite view to find narrow bridges. Lots of streets and bridges are only 1-2 meters wide in the pre-1460 style street pattern. The ones right NE and ENE of the center are interesting, but so are some ESE further away.

You can see the agricultural chinampas from a trajinera at Embarcadero Cuemanco (on the freeway) or Embarcadero Fernando Celaya just west of the center. The embarcaderos the big signs direct you to right off the light rail are very commercial, expensive, and oriented to the plant nurseries and mariachi bands. That's a Mexican cultural experience, too, but not as deeply Xochimilquense.

If you're feeling very adventurous, rent canoes at the Michmani Ecotourist center, 50 m south of Embarcadero Cuemanco. Set aside 2-4 hours at least to paddle through the canals. Bring sunscreen, water, and a map (inkjet printed or smartphone is fine). Be sure to ask to see the endangered ajolotls being raised at Michmani.

Also, the centro Xochimilco is interesting and the Dolores Olmedo is essential if you care about Frida and Diego.

By 1519 it was the largest city in the world.

Where did you get that idea?

Attila the Hun prompted the founding of Venice, back in the 5th Century. Attila was eagerly pillaging and slaughtering everything on the mainland-- so those awful marshy islands quickly became very attractive to the mainland locals; the Huns operated mostly on horseback, didn't like boat building, and had much easier targets elsewhere.

Venice thus turned out to be safe from Attila and most of the rest of the endless European conflicts that followed-- that was its primary attribute for economic progress and longevity. The random narrow channels between the original marshy islands became the now famous canals of Venice.

Geography was the key to Venice, but from a military defensive viewpoint most of all.

One of the reasons is that canals facilitate the transmission of diseases.

Diseases where linked to water and then microbes just 160 years ago =(

Of course, but that didn't stop people from dying whenever they decided to settle in a malarial swamp.

One of my pet peves is the notion that people were incredible idiots in the past, unable to make the most basic logical connections. Marcus Terentius Varro said the follow, in ancient Rome:

Especial care should be taken, in locating the steading, to place it at the foot of a wooded hill, where there are broad pastures, and so as to be exposed to the most healthful winds that blow in the region. A steading facing the east has the best situation, as it has the shade in summer and the sun in winter. If you are forced to build on the bank of a river, be careful not to let the steading face the river, as it will be extremely cold in winter, and unwholesome in summer. 2 Precautions must also be taken in the neighbourhood of swamps, both for the reasons given, and because there are bred certain minute creatures which cannot be seen by the eyes, which float in the air and enter the body through the mouth and nose and there cause serious diseases.” “What can I do,” asked Fundanius, “to prevent disease if I should inherit a farm of that kind?” “Even I can answer that question,” replied Agrius; “sell it for the highest cash price; or if you can’t sell it, abandon it.”

Of course people made the connection between settling in a swamp and the spread of diseases. Before germ theory gained wide acceptance you had the notion of there being "bad air" around swamps which caused disease.

Awesome. The germ theory of disease before the birth of Christ. Imagine what Varro might have accomplished if he had had a microscope.

Point taken, but that was the Roman elite discourse. How about the people that actually decide where to build new settlements? If you look at 2014 slums, people still hasn't got the "bar air" idea.

Moreover people knew that bad water spreads disease even if they did not know the exact mechanism

A lot of coastlines are either unprotected from major waves coming from the open ocean or are indistinct swamp and mudflats where it's hard to get a big ship close to dry land. Harbors where you have a sharp transition from deep water to dry land, such as San Francisco and Manhattan, are fairly rare and thus valuable.

You can't build Venice just anywhere; specific geographical features are needed. Further, you wouldn't want to build Venice unless building on more solid land wasn't possible. It is expensive to build, expensive and inconvenient to live there, and faces constant flooding problems. Is this really a mystery?

This is the place to post if you are confused about TC's post. As originally stated: "That is, why are there so few cities with multiple water connections between different independent nodes?" - I don't see that there are few cities. In fact, historians say every major city is on or near the ocean or water, so in fact there are multiple water connections. (This is because water transport is cheap compared to overland traffic, malaria notwithstanding). Perhaps TC is commenting on the fact that not all cities are as interconnected as they can be, like a star topology, but I think network theory says you don't need to have every city connected to each other for the free flow of traffic between nodes, just a dominant few nodes need interconnection.

Edo (now Tokyo) very closely resembled Venice with canals everywhere. You can see this in wood block prints and art from the era, e.g.:



Most of the canals are now gone because of economic development. The Tokyo Olympics were the last straw - lots of new roads build right on top of old canals to improve urban infrastructure.

I suspect the lack of development is a big part of the story with Venice - the city didn't become a modern metropolis, so the canals were left untouched.

+1. It is also interesting to look at the effect on the surrounding countryside and their access to the waterways of the city. Before the combustion engine made trade by road feasible on a large scale in Tokyo, neighboring Chiba prefecture was far richer than Kanagawa Prefecture on the other side of Tokyo due to easy access to Tokyo's waterways. As International trade became more important Tokyo centralized its port trading functions and more or less cut Chiba off. Now Kanagawa is richer.

The role of the surrounding countryside is the difference between Venice and a village on stilts in a lagoon.

Coastal cities need water, not too much and not too little.

Rome's ancient seaport of Ostia Antica is now 3km from the sea, while Cleopatra's palace lies below this same sea.

Venice in Southern California, just south of Santa Monica, was built in 1905 around canals. For various reasons, it turned out to be a disaster, the Slum by the Sea, that took ridiculously long to gentrify.

Primarily due to something more valuable then having scenic canals - oil. Venice, CA of 1925 is a forest of oil derricks - with the resulting pollution and foul air.
Once the derricks left and LA ran out of sprawl, it could be redeveloped.
But the canals were always artificial; just an attempt to look like Venice, not to actually travel by boat through necessity.

By the way, just south of Venice, CA is Marina Del Rey, Los Angeles's main sailboat harbor. It opened in 1965 and has proven a more successful attempt at artificially integrating water and land. Interestingly, although it was designed with many channels, I don't believe it has any islands (which saves on the cost of bridges -- you can get anywhere on land in Marina Del Rey by car, but it's often a long drive around to get there). Unlike Venice and like Duluth, it has only one opening to the open water: a long channel with a perpendicular breakwater to keep the big Pacific breakers from rolling in and swamping the boats.


I would imagine that more recent developments tend to build islands because people like islands for interesting psychological reasons and thus will pay more for a house on an island than on the mainland. But Marina Del Rey was a product of a period in American history when design tended to be pragmatic and kind of cheap. I suspect its design represents a basic practical logic, even if its lacking in the romance of having many islands.

By the way, if you were designing a real estate development, I bet you could really charge a lot of money for an island in a lake on an island. It's not very practical, but it sounds cool, especially if you are nine years old.

"There is a long history of stilt houses on water": some crannogs may have had that form.

Sowing needles or sewing needles? I am not familiar with the first, but they sound intriguing.

I sometimes don't get these posts.

People only build a city on the islands in the Venetian lagoon when the war between the Byzantines and first the Goths and then the Lombards made the adjoining areas of the Italian mainland too dangerous. There had always been cities in what is now northeastern Italy, but during the period Venice was founded they kept getting destroyed. There happened to be a shallow lagoon with some islands that people could move to. They had been ignored in the past precisely because it otherwise wasn't a very good place to put a city.

The leaders of Venice they brilliantly exploited their position of being not quite part of the Byzantine Empire and not quite part of Catholic medieval western Europe to first dominate trade between western Europe and the eastern Mediterranean, and then build an small empire. Even then, there were close calls and alot of luck. Its pretty obvious why the history of Venice is unique.

The history of San Francisco also shows what a big part accidents play in the growth of a city. It happened that the gold rush occurred at the same time in California during the few decades that sailing technology made it desirable to put the port on the northeastern side of the mountainous peninsula, just part the Golden Gate, instead of normally more logical sites on the East Bay.

It also mattered that the Med, and therefore the Adriatic, is almost tideless.

It's also stormless; Venice would not fare well in a hurricane.

Unrelatedly, I would like to share my appreciation for the line "But if you are a Marsh Arab wishing to trade some rice for some coriander with your neighbor, and the next day lend a sowing needle in return for some gossip," which made me literally burst out laughing.

Duluth, Minnesota and adjoining Superior, Wisconsin owe their existence to the transfer of two bulk commodities, iron ore (now taconite pellets) and grain. In reality most of the port facility is located in Superior. The only other water connection to the two towns is the St. Louis River, unnavigable just a few miles up stream.

Also, the water around Duluth has a significant tendency to be in solid form for a long stretch every year.

I'm surprised not to see Saint Petersburg mentioned. Russians, at least, always refer to it as the "Venice of the North", it has many canals, some rivers, and a few openings to the Bay of Finland.

Disease, as mentioned above, also it's easy to transport bulk goods by water, but not as easy for humans to travel. There are obvious reasons for preferring dry land.
I would guess that canal cities only develop when people have no other choice of where to build. Like getting attacked by goths, or because an ideal trading site where two giant rivers meet is inopportunely located in a huge marsh.

Don't forget the Seven Bridges of Königsberg.

Let's not forget Venice, California, which is possibly an improvement on the original (speculative)

So far as I know, Venice's location (marshland) was chosen to better isolate them from the advancing Germanic invaders. Anything they did with their locale, afterwards, may have been simply "damage control." Trade made Venice, not canals, etc., and being in a sensible general location for trading is more important, and as some have pointed out, there is better infrastructure.

There are aesthetic benefits to this type of urban layout -- the Palm Islands in the UAE demonstrates this -- but this development is more about consumption rather than (urban) investment..

Venice has water connections between two significant independent nodes: boats/barges on the Po river, and deepwater shipping in the Adriatic. This is how pretty much every great city on the planet is situated - one navigable river meets one sea or ocean. If you find a major city inland, it's usually at the junction between a river and a major tributary, or near the head of navigation of a river. But basic topology makes it rare to have more than one river meet the sea, or two rivers meet each other, in the same place.

That Venice uses canals for intraurban transport seems to me of little significance. The mouth of the Po is marshy enough that anyplace you could reasonably chose to put the necessary port city you'd find it cheaper to dredge canals and build lots of little bridges than to provide solid ground for streets. Fine, and as you note that's true of a few other places. But I'm not getting the part where this provides greater economic opportunity. If a Venetian wants to trade with someone outside Venice, he goes to the port district - which does not seem to be any larger or spread-out than that of any other major city. The long-range shipping is not coming to the doorsteps of every Venetian, or even every Venetian business. And if a Venetian wants to trade with another Venetian, he walks or takes a gondola to the other guy's shop. In most cities, that would be "walk or take a horsecart", so what's the difference?

"the (now largely defunct) Marsh Arabs from southern Iraq"

Yeah, as if they mysteriously disappeared, or all decided to leave, or something. It's not like they were crushed by Saddam after we encouraged them to revolt and then abandoned them... https://en.wikipedia.org/wiki/Marsh_Arabs#1991.E2.80.932003

Probably had more to do with Saddam draining the marshes in perfect Soviet style to try to turn the area into an agricultural basin.

Over the 19th Century, land transport improved far more than water modes. Putting an engine on a wheeled cart was quite a revolution.

The engine on the wheeled cart isn't going anywhere without decent roads.

Between Oakland and San Francisco there is a brand new bridge, a reasonably good subway tunnel and a ferry.

Hardly anyone takes the ferry.

Boats are great for moving cargo over long distances but cheap oil and high quality road/rail networks make shipping people over short distances a largely terrestrial activity.

The advantages of using water-based people transportation disappeared by about 1925 and are unlikely to return any time soon. Given that the overwhelming majority of urban growth has occurred over the past century, it shouldn't surprise us that cities are rapidly filling in their canals to build roads.

I travel back and forth to Boston and its South Shore by high speed ferry, which services a lot of people. The all in economics are far better than the train alternative, which was and is heavily subsidized by the taxpayers. Another advantage is that the ferry is also far more dependable in the winter.

I feel the major reason is disaster - Venice seems to be in a pretty good spot in the Adriatic. But if you live on any coastline that deals with Tsunamis, being right out there in the open water means eventually the odds will run out and the city will be destroyed (and the resulting survivors will decide to move inland and avoid the wrath of the angry deity that sent the oceans against their proud city).
Also, even aside from Tsunamis, large winter storms or other types of floods are the danger. It's not a risk free proposition.
A modern city (or development) adds inland canals for recreation, not trade - a place to tie up a boat as a form of status symbol, with the use of the boat during leisure time. Individual commuting and trade is performed on the highways, bulk trade on through the primary ports, roads, and highways due to efficiencies.
Even Venice itself is more of a living museum - the major port for bulk items is by the mainland, and most travelers arrive to Venice by train from the mainland (except for some tourist cruises).

There was a precursor to Venice, Ravenna, a city in the marshes of the Po where the Romans had a naval base. The Western Roman government relocated there because it was the most defensible place in Italy against barbarian raids, so it became the last important city in Italy before the Dark Ages, despite being in a really unhealthy climate.

The Lombards finally captured and destroyed Ravenna in the 8th century, and the marshes silted up so the city wound up well inland. The rise of Venice really got going only after Ravenna was destroyed.

I'd like to add more definition to the term "Venice-Like-Cities" than the number of "water connections between nodes". There are huge differences in how those connections came into being.

Starting a settlement on an (archipilago/river) island is something that used water for protection, e.g. Stockholm and Paris. That Paris is not coming into mind as a Venice-look-a-like but Stockholm does highlights the superficiality of the term - both cities grew beyond their core island. It just happens that Stockholm had many islands around it, paris didn't, making Stockholm a 'Venice of the north'.

Yes, a connection to water roads was always seeked for a new settlement but not necessary. It helped with access to markets, to energy and ... well water. Before the arrival of steam power, water connections were by far the most efficient form of transportation. But only with the arrival of mercantilism efforts were made to actually shape the geography to create more "water connections between nodes".

The two main factors benefiting such construction developments were market-potential and the affordability. Along the continental coast of the north sea, you will find many "Venices of the north" since creating a water road just afforded you to dig a meter deep. If your land is already beneath sea-level, affordability was even less a restraining factor since water roads are even cheaper to build there and were also benefitting flood-safety of your whole low-lying lands. Thus, almost every town in the Netherlands is "venice-like". I guess the same could be true for asian low-lands as well.

You can see the effect of channel-affordability in reverse in cities like Hamburg, where with the arrival of motorization many of the dug-out channels were filled up. Until the end of the 19th-century almost every building of the Hamburg core-town had direct access to a water road.

So what makes a city truly venice like? The highly sophisticated effort to create a settlement specifically for the purpose to make use of as many water roads as possible. And I might add the means of achieving this are important as well: Creating Islands instead of digging out channels and letting them fill up with water.

The purpose of increasing goods-accessibility of households by water-roads back then is comparable to todays broadband-internet-infrastructure projects. Making island or water road developments in Miami-Beach or Lingang-City less Venice-like because they were solely created out of real estate purposes.

I suppose the real venice-like standards are only met by Mexico City and St. Petersburg? At least those cities are more Venice-like than Hamburg (called "venice of the north" because it has more bridges than...) or Amsterdam (also called...). And probably any of the european low-land cities is more Venice-like than Stockholm.

The United States, the United Kingdom, their colonies, and quite a few other European colonies derived significant revenues from customs duties. Custom's duties were the dominant source of federal revenue prior to the Civil War in the U.S. and remain the dominant source of public revenue in many island nations and financial centers. It is hard to collect customs duties in Venice. It is easy to connect customs duties in port cities with only a few points of entry. Therefore, there is a state incentive to regulate land use to facilitate custom's collection.

You also need, if you want a Venice, a relatively sheltered main body of water like the Adriatic. If you face a big ocean, the likelihood of a storm surge is too great to invest in developing it with infrastructure to close to the waterline. For example, recent paleoclimate data from Florida reveals that it has experienced a 16 foot storm surge from hurricanes on average ever 40 years out of the last thousand years or so (the last 160 years has been ten times more mild). Storm surges destroy the infrastructure that makes a Venice worth keeping. Look what one bad storm (Katrina) did to New Orleans.

I've been to Miami only once in the past 50 years, but was struck by the large number of canals around there, not just around Miami Beach but throughout the area. According to the Florida Fish and Wildlife Commission, they were built in the early 1900s for drainage and flood control, which is not a reason that I see cited for other cities' canals. Perhaps being in a wet, flat, low-lying area means that you've got so much water that storm drains and the like are inadequate so you use canals to channel water instead.

Florida golf courses have a lot of water hazards for similar reasons. If you have a bunch of swamp land, you dig out lakes and pile the dirt up to make the fairways.

Suzhou is the Venice of the East.

I had always thought Bangkok was the Venice of the East, but there does appear to be some confusion.


Let's not forget Bruges and Ghent in Belgium.

Comments for this post are closed