What is a city, really?

In 2019, I was asked, at the last minute, to attend the Smart City Expo (the person who was meant to go had a visa problem, I was told). I wasn’t very interested in smart cities, and if things had gone to plan, I never would have known the conference existed.

I sat through every panel and meeting from the opening to the closing; I looked at the solutions on the exhibition floor, took notes and took photographs. Back home, I thought, someone might ask what I had learned, and I would put together a presentation.

Nobody asked anything.

This series grew out of the questions in that presentation — the one nobody asked for, and I never made.

Of the solutions on the floor as a whole, I had noted (translating from my own Turkish): “it looks like an entertaining Lego puzzle area where the same architectural boxes are put together for different problems…, and the problems that will surface at implementation have been swept under the carpet.”

Smart devices at the edge; a network like a nervous system, connecting them securely to one another and to shared centers; and in those centers, a mind running an endless loop at different time scales: notice, decide, respond.

I had not seen that loop for the first time at the fair; I had already been thinking in those terms in some of the solutions I had designed myself: sense (notice), decide, act. I suppose a person sees more easily what he already knows — knowing, seeing, being.

So I started from the first step: “notice.” Notice what? The city.

I wrote this first essay purely out of curiosity, but the answer I give to what a city is may well shape how I look at the nine questions that follow.

So, “what is a city?”

Trying to learn the answer took me from Göbekli Tepe to Uruk and Çatalhöyük, into the anthropology literature, and into the notes of the many archaeologists and sociologists who have thought about it.

The tidiest answer I found came from a sociologist. In 1938, Louis Wirth defined the city in four words: large, dense, permanent and heterogeneous.¹

Those four words give me a good starting point: they tell me what to look at when I examine a settlement. They matter for the smart city too; if we want to make a city smart, we first need to know which of its properties we can measure. Counting the population is easy; is measuring how much people who do not know one another depend on one another just as easy?

Large?

The first answer that comes to mind is population. A village is small, a town medium, a city large.

In most of the world, population is the first criterion for ranking settlements. But where is the line? In one country a few thousand people count as a city; in another, fifty thousand. There is no agreed number.²

Çatalhöyük, in Konya, was one of the largest settlements of its time; Scientific American calls it “the metropolis of its day.”³ How many people lived there at its peak is still debated. Yet archaeologists cannot call it a city; they call it a “proto-city”: like a dozen villages grown together (this reminds me of one of Turkey’s cities, but I won’t say which) :-)

So size alone is not enough.

Dense?

Çatalhöyük was dense. The houses leaned against one another and were entered from the roof; there was no street, no square, no temple, no ruler’s house; the houses were all more or less the same.⁴

That archaeologists hesitate to call Çatalhöyük a “city” shows that crowding and closeness do not quite amount to a city either.

Permanent?

Permanent does not mean people coming and going; it means living there and staying — all year round, for generations. Göbekli Tepe puts this word to the test.

Eleven thousand years ago, on a hill near today’s Şanlıurfa, a great many people gathered. They raised stone pillars to five meters tall and carved foxes, wild boars and vultures onto them. There was as yet no farming, no domestic animals, no pottery, no writing.

These people were hunter-gatherers. They are said to have returned to the same place again and again for about fifteen hundred years.⁵ Communities came from an area two hundred kilometers across to work together and eat together.⁶ Gazelle bones suggest the gatherings were seasonal.⁶

It was not permanent — at least most of those who came did not stay. So why put so much labor into a place where they weren’t going to stay?

So why did people come together?

The most common account: first came farming; farming produced a surplus; the surplus fed a crowd; and the crowd built the city.

In 1950, Gordon Childe turned this into a list: surplus, full-time specialists, monumental buildings, a ruling class, writing, long-distance trade… If a place had these, it was a city.⁷

The story is neat, but it has cracks, too. The anthropologist Marshall Sahlins argued that hunter-gatherers had more leisure than farmers.⁸

So why did people move to the harder life of farming? A growing view holds that many settlements were founded before agriculture.⁹

According to Klaus Schmidt, who led the excavation at Göbekli Tepe: “First came the temple, then the city.”¹⁰ On this view, people may have gathered not for food but for a shared belief — and then, to feed themselves, begun to sow the wild wheat nearby.

Agriculture may have been not the cause of the gathering but its consequence.

Today some researchers go further still and propose that urbanity first emerged in sacred places, in rituals and festivals.¹⁰ This reading is debated too. Lee Clare, who leads the Göbekli Tepe excavation today, says the evidence for at least a semi-sedentary population is growing.¹¹

According to the German Archaeological Institute, Göbekli Tepe is not an isolated exception; together with neighbors such as Karahan Tepe, it is part of the “Taş Tepeler” — Stone Hills — world.⁶

So Göbekli Tepe had the muscles of a city: labor that organized many people, a shared meaning, a continuity that lasted fifteen hundred years. To my mind, this was the closest thing to a city that nomadic hunter-gatherers knew.

Is a temple required?

So, as Göbekli Tepe suggests, must a city have belief at its center?

Egypt’s first walled towns clustered around temples; but at Harappa and Mohenjo-daro in the Indus Valley — big cities with planned streets, drainage and long-distance trade — archaeologists found no palace and no royal tomb, nor any building that can firmly be called a “temple.”¹²

These cities turned, it seems, not around a god but around trade and shared rules.

So a temple is not required either.

Heterogeneous?

For Wirth, heterogeneous means people who do not know one another, and who do different work, living side by side.

The first time we went to a friend’s village, I was surprised that everyone seemed to know who I was; in the city we share the street with strangers, so it seems natural that we need rules, institutions, and writing.

Those who came to Göbekli Tepe were most likely communities who knew one another; they gathered and dispersed, and few strangers were among them.

One item on Childe’s list matters here: full-time specialists. People who do not grow their own food — artisans, scribes, priests.⁷ They depend on what others produce, and others depend on their work. Çatalhöyük had a crowd, but there is little trace of full-time specialists; every household did its own work.

Uruk, in Mesopotamia, about five thousand years ago, had all four: tens of thousands of people in a walled, crowded city; a settlement that lasted thousands of years; artisans and scribes who did not grow their own food. The earliest known writing also appeared there — to keep accounts.¹³

Archaeologists call Uruk a city.

What I take from all this:

A city is a place where people who do not know one another and do not produce their own food depend on one another. The crowd is a precondition for that; it is not the definition.

So what is a city?

Size alone is not enough, nor is closeness; a temple is not required. What remains is this: might a city be, more than a physical place, a pattern of relationships?

Of Wirth’s four concepts, this may be the most important: “heterogeneous.” An order in which people who do not know one another meet every day, depend on one another, and invent rules, institutions, and writing to sustain that bond?

The stone buildings look like the vessel of that order, not the order itself. That is why I hesitate when I look at Göbekli Tepe: some pieces of the vessel are missing, but part of what it held — the shared labor of the people who gathered there, the meaning they made together — can still be read.

The Lego boxes I saw in Barcelona counted streets and vehicles; can they count relationships too? What can the “notice” step see?

Is a city a machine that can be tuned, or a living organism that answers every intervention in its own way?

I think that is the next essay’s question: What kind of problem is a city?