How big is the smallest city?

a cityscape at night with buildings reflected in water

In a thoughtful CfAS blog post over the holidays, Hagit Nol discussed some of the challenges of defining the concept of a city in an archaeological context. Reading that post really got me thinking, so I thought I’d share my thoughts about what makes a city here.

Nol mentions an important distinction drawn by Michael Smith and Jose Lobo (2019) between sociological and functional definitions of cities. Roughly speaking, that’s the difference between cities as places with lots of people and cities as places with lots of purpose.

My own somewhat ecumenist view is that both offer a grain of truth. To see why, it’s probably easiest to start with why a functional definition by itself won’t do. Mainly, that’s because any set of functions you can think of to define a city would almost certainly apply to urban concentrations that are not cities, too, like towns, villages, and boroughs. That, or they would include functions that only accidentally differentiate cities from those other urban concentrations – meaning, they don’t really tell us what a city is, just what it happens to be.

Note that the claim here is not that no set of functional roles is necessary for an urban concentration to count as a city, just that no set of functional roles is sufficient for being a city. I think to get to sufficiency, you need function and size. That’s because cities are just big towns.

This, of course, raises the obvious question: how big does a town have to get before it becomes a city? Here, I think, is where the definitional challenge of the city concept takes a more precise form.

Consider Tokyo, Japan, with its population hovering around 14 million. It’s as obvious a candidate for a city if there ever was one, but what about my hometown, Conway, Arkansas? In the eighties, Conway was a quiet town of about 20,000 people. It has since tripled in size. Is it a city now? My intuition says no. At least, it’s not a city in the way that Tokyo is a city, but what if it keeps growing? Eventually, it will reach a point where no reasonable person could deny its membership in the conference of cities. At the same time, there’s no obvious point during its time of growth that you can zero in on and confidently say, “Right here, this is the exact moment where it became a city,” even though the Law of Excluded Middle demands it of us.

This sort of thought experiment, an instance of what philosophers refer to as the Sorites Paradox, illustrates an important fact about the ‘city’ concept, namely, that it is vague. It has fuzzy boundaries. Asking how big a town has to get to become a city is like asking how many grains of sand it takes to make a heap.

Philosophers (probably for reasons of job security) see in the Sorites Paradox a fountain of deep metaphysical and epistemological implications, but for me it’s just a reminder that language evolves. In this case, our concept of a city evolved to have just as much precision as was needed for a person to say, “That city over there,” and have a reasonable shot of being understood. If our meaning was unclear in such cases, we could always add more information to narrow in on our reference, like “That city over there, the one with the giant space needle,” without having to engage in a lengthy philosophical debate about the nature of cities.

Thinking about the city question this way also helps us to understand why people can reasonably disagree about whether an urban concentration should count as a city. People will typically agree that, for example, “water is wet” and “grass is green,” but underlying that agreement is a set of unspoken expectations about what should count as normal conditions.

For water being wet, those normal conditions would include temperatures in excess of the freezing point of water. For grass being green, they would include a certain amount of precipitation. For the claim that cities are just big towns, however, our only anchor points are smaller urban concentrations, so it seems only natural that people who grew up in a world with, say, 500 million people should have a different threshold for city-hood than people who grew up in a world with 8 billion people. Or why a kid from Podunk, Arkansas, might be gob-smacked by the sprawling metropolis that is Dubuque, Iowa (apologies if you are from Dubuque).

The problem for us as scientists studying cities is that the vagueness of the word means there’s no “fact of the matter” about whether edge cases should go one way or another, so no amount of “archaeological evidence” could ever really settle the matter. Nor could we devise a special sort of city-detection device to answer the question once and for all, not at least until we agree on some more precise criteria anyway.

Alternatively, we could just make up a term, say a “principal city,” and define it as, for example, a city-like “incorporated place with … at least 10,000” people in it, like the US Office of Management and Budget does. Most of the questions we would want to pose about cities, we could then ask more precisely about principal cities.

However we decide on this though, for me, the critical element is consistency, the consistent application of a criteria (whatever it is), as our errors, if they ever come to light, will all go in the same direction and as a consequence be more amenable to correction. Consistency would also ensure that the data are comparable. I would even argue that consistency (or at least some level of consistency) is necessary for something to count as data at all! There are also the practical benefits of consistency, like making it easier to synthesize data into larger regional datasets, conduct meta-analyses, and replicate results.

Of course, the concept of a principal city has the full might of the US federal government behind it (at least until the end of 2024…). But there’s no comparable political entity within the community of archaeological science. Still, assuming some of us can reach some consensus around a set of standards, what we will do as archaeologists to encourage their adoption? When it comes to advancing the archaeological study of cities, this is where the rubber really meets the road.

Like this article?

Share on Facebook
Share on Twitter
Share on Linkdin
Share on Pinterest

1 thought on “How big is the smallest city?”

  1. “Grass is green” only in those societies whose language divides the spectrum so as to have a term for that section that WE call “green”. In many languages, such as Blackfoot, the spectrum divides to lump our blue and green into one color, what anthropologists call “grue”. Its central focus area is what we call aquamarine. Language tends to guide perception, a sobering thought.


Leave a Comment